US6388947B1  Multicrosswell profile 3D imaging and method  Google Patents
Multicrosswell profile 3D imaging and method Download PDFInfo
 Publication number
 US6388947B1 US6388947B1 US09152935 US15293598A US6388947B1 US 6388947 B1 US6388947 B1 US 6388947B1 US 09152935 US09152935 US 09152935 US 15293598 A US15293598 A US 15293598A US 6388947 B1 US6388947 B1 US 6388947B1
 Authority
 US
 Grant status
 Grant
 Patent type
 Prior art keywords
 model
 data
 seismic
 crosswell
 method
 Prior art date
 Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
 Active
Links
Images
Classifications

 G—PHYSICS
 G01—MEASURING; TESTING
 G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS
 G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
 G01V1/40—Seismology; Seismic or acoustic prospecting or detecting specially adapted for welllogging
 G01V1/42—Seismology; Seismic or acoustic prospecting or detecting specially adapted for welllogging using generators in one well and receivers elsewhere or vice versa
Abstract
Description
This invention was made with Government support under STTR GRANT No. DEFG0294ER86019 awarded by the Department of Energy. The Government has certain rights in this invention.
The present invention relates generally to crosswell seismic mapping or imaging and more particularly to a method for producing high quality three dimensional mapping or imaging using at least one crosswell data profile in combination with other downhole derived data which may include one or more other crosswell data profiles.
In the field of geophysics, the knowledge of the subsurface structure of the ground is useful, for example, in the extraction of mineral resources such as oil and natural gas. In the past, a number of different seismic imaging methods, including tomographic methods and reflectance imaging methods, have been implemented with the goal of rendering images which impart such knowledge of the subsurface geologic structure. With regard to an oil or gas reservoir, the availability of an accurate prior art seismic image, as interpreted by a geoscientist reasonably versed in its use, can provide the capability to extract additional production from a reservoir than would otherwise be possible. Thus, in view of the earth's limited resources and the expense encountered in recovery of such resources, prior art seismic imaging has become increasingly important.
One area of particular importance with regard to prior art seismic methods is that of crosswell seismic imaging. Conventional crosswell seismic imaging typically utilizes a pair of boreholes in proximity to the reservoir. In the first of these boreholes, a seismic source is deployed for emitting seismic energy into the region of interest, often as a swept frequency signal (chirp) which covers a predetermined frequency range. The source is selectively moved between a series of positions within the first borehole at predetermined times. The seismic energy passes through and around the subterranean region of interest to the second one of the pair of boreholes. A receiver array is typically deployed within the second borehole and, like the seismic source, the receiver array is moveable between a series of positions within the second borehole. It should be appreciated that the subterranean region of interest may comprise a zone which is at a known or estimated depth range below the surface. In this case, the receiver array and source positions are adjusted accordingly such that the positions are spaced across the zone of interest rather than extending all the way to the surface. By transmitting from each source position in the first borehole while receiving data at each and every receiver array position in the second borehole, a seismic crosswell dataset is generated. Crosswell seismic datasets typically offer advantages in resolution over surface seismic datasets which are more commonly practiced.
After having generated a seismic dataset, the task of using the dataset to produce a crosswell image or tomogram may be undertaken. In this regard, it should be appreciated that the crosswell seismic dataset comprises a large, complex set of information which is rich in detail relating to the geologic structure and material properties of the subterranean region of interest. Due to this complexity, certain known types of data such as, for example, direct arrival traveltimes are generally obtained from the profile and utilized iteratively in a mathematical model of the subterranean region to establish the geologic structure and material properties within some approximation. While a number of effective methods for generating models have been developed in the prior art, it should be appreciated that most conventional crosswell imaging is limited to producing a two dimensional tomogram in the plane defined between two generally vertically extending boreholes. The usefulness of even a two dimensional high resolution tomogram depicting a “slice” of the subterranean region between a pair of boreholes is not diminished by the teachings and discoveries disclosed herein. However, these discoveries in conjunction with other recent developments in the prior art are thought to provide sweeping improvements.
One recent development relates to the practical acquisition of individual crosswell datasets. In the past, data was collected between a pair of boreholes in the manner described above. The collection of even one crosswell dataset was prohibitively expensive, at least in part due to relatively inefficient methods. More recently, however, improvements in data acquisition technology have been made which have dramatically increased data collection efficiency between a pair of boreholes. Such improvements include, for example, stronger sources, shooting “onthefly” and the use of multilevel receiver strings. Based on improved collection efficiency alone, the amount of data which may be collected in a single day has increased by a factor of at least ten. Such a dramatic improvement makes crosswell seismic even more attractive and opens the potential for crosswell surveys over large areas of existing and prospective reservoirs. Of course, the advantages of crosswell seismic such as, for example, high resolution (on the order of well log resolution) remain attendant to its use. However, there remains a need for a corresponding improvement in the all important image generating process for dealing in an effective way with this vast increase in the amount of available data, particularly in the instance where a number of crosswell datasets are available for a single field or reservoir. This need as well as other needs for improvement will be discussed in further detail below.
A number of crosswell datasets may be available for a particular reservoir or region for two reasons. First, the datasets may be obtained in the conventional manner between pairs of boreholes across the region. Second, using more recent data acquisition technology, crosswell data surveys may now be conducted across a region penetrated by a plurality of boreholes by deploying a source in one of the boreholes and deploying receiver strings in each of the remaining boreholes so as to simultaneously record the output of the source to at once provide a plurality of datasets. In this regard, one of the more telling limitations of current crosswell technology resides in the fact that conventional crosswell technology provides only the capability to use each dataset apart from the other datasets to generate a two dimensional image. That is, the geoscientist is presented with a plurality of independent, two dimensional images, one for each dataset. In most practical cases, two dimensional planar images are an approximate depiction of the true complex, three dimensional earth structure. Unfortunately, a more detailed limitation has been observed in this regard to the extent that intersecting planar images produced between different pairs of boreholes in the same region commonly exhibit inconsistencies at intersection points of the two images so as to present a dilemma as to the actual structure. Presently, the only solution to this dilemma has been found in questionably effective techniques such as interpolation. Applicants are not aware of any effective prior art technique for producing a uniform image across a region in view of a plurality of independently generated crosswell seismic planar images, irrespective of the fact that the datasets resulting in the images may even have been obtained simultaneously.
Still another and practically significant limitation of conventional crosswell seismic imaging is found in situations where boreholes from which a crosswell dataset is generated exhibit significant deviation from the vertical direction. One inherent assumption of conventional crosswell imaging is that of producing an image of the plane defined between a pair of vertically extending boreholes. Unfortunately, this assumption is problematic in instances of significant borehole deviation. In the past, correction schemes have been introduced in an attempt to compensate for such deviation. However, there remains a need for a more effective solution. Similarly, conventional crosswell imaging assumes generally two dimensional earth structure with limited geological complexity. Where these assumptions are not valid, the planar images will be erroneous. Therefore, the need remains for improvement in capability for use in regions which exhibit complex geologic structure including, for example, high velocity contrasts and/or significant out of plane dip.
Prior art crosswell seismic imaging has drawn, at least in part, from the arts of surface seismic and surface to borehole seismic techniques. Consistent with this background, most practical and proposed methods for imaging from crosswell data are based on conventional, pixelized, model representation. Unfortunately, pixelized model representations, whilst highly flexible in the limit, generally rely on numerical methods for processing and become computationally inefficient and even impractical as model resolution and complexity are increased. Specific methods with advantages for crosswell imaging, particularly tomographic and inversion methods, as well as schemes for analytical representation of seismic model parameters, are well described in the literature, as listed below:
Phillips, W. S. and Fehler, M. C., 1991, Traveltime tomography: A comparison of popular methods: Geophysics, 56, no. 10, 16391649.
Guiziou, J. L., Mallet, J. L. and Madariaga, R., 1996, 3D seismic reflection tomography on top of the GOCAD depth modeler: 61, 5, no. 1, 14991510.
Chiu, S. K. L., and Stewart, R. R., 1987, Tomographic determination of threedimensional seismic velocity structure using well logs, vertical seismic profiles, and surface seismic data, Geophysics, 52, 8, 10851098.
Dynes and Lytle, 1979, Computerized geophysical tomography, Proc IEE, 67,10651073.
Scales, J., 1987, Tomographic inversion via the conjugate gradient method, Geophysics, 52,179185.
Scales, J., Doherty, P., and Gerztenkorn, A., 1990, Regularisation of nonlinear inverse problems: imaging the near surface weathering layer: Inverse Prob, 6, 115131.
Gerskentorn, A., and Scales, J., 1987, Smooting seismic tomograms with alphatrimmed means: Geophys. J. R. Astron. Soc., 91, 6772.
Meyerholtz, K. A., Pavlis, G. L., and Szpanowski, S. A., 1989, Convolutional quelling in seismic tomography, Geophysics, 54, 570580.
Therefore, direct comparisons will be made at appropriate points below between the present invention and the ‘stateoftheart’ in crosswell traveltime tomography serving to point out significant differences which ultimately relate to the advantages attendant to the use of the present invention. The principal methods of prior art crosswell traveltime tomography are outlined in Phillips and Fehler (1991) and include:
Damped least squares or iterative backprojection (Dynes and Lytle, 1979).
Damped least squares with smoothing (Gersztenkorn and Scales, 1987).
Damped least squares with convolutional quelling (Meyerholz, et al, 1989).
Regularized inversion (Scales, et al, 1990).
Iteratively reweighted least squares (Scales, et al, 1988).
It is noted for later reference that all of these principal methods utilize iterative backprojection through a slowness model, and therefore are subject to the constraints and limitations of the crosswell aperture (Rector and Washbourne, 1994) and the resolution limits imposed by ray tomography (Williamson and Worthington, 1984).
Attention is now directed to FIGS. 13 for purposes of providing an actual example illustrating still other limitations of prior art crosswell seismic tomography not previously mentioned. Specifically, FIG. 1 illustrates a region 10 which forms part of the McElroy field in the West Texas Permian Basin. A crosswell data survey was performed in 1995 for use in mapping the region. The reservoir depicted in region 10 is a structural/stratigraphic trap about 2850 feet deep and 150 feet thick. The area was a CO2 flood pilot wherein the objective of the survey was to evaluate hybrid water/CO2 injection. Seismic velocity variability in the region is extreme, about 50%, ranging from 14,000 to 23,000 ft/s. A series of four crosswell data profiles was obtained using four relatively vertical wells indicated by the reference numbers 12 ad. Arrows 14 ad indicate the various profiles wherein the arrows are directed from the source well towards the receiving well. It should be noted that constrictions in wells 12 b (1202) and 12 d (661) prevented logging to maximum depth. As will be seen, such constrictions have significant effects on the two dimensional coverage of prior art velocity tomograms.
Referring to FIGS. 2 and 3, the well geometry of FIG. 1 has been “unwrapped” in a way which facilitates simultaneous viewing and comparison of all four prior art tomograms available from the crosswell survey wherein the unwrapped view is diagrammatically shown in FIG. 2 and a grayscale representation of the actual tomograms is given in FIG. 3. In this regard, one of ordinary skill in the art will appreciate that tomograms are normally depicted in color such that different colors represent velocity variations. Unfortunately, the present forum does not provide for a color presentation. However, the grayscale reproduction of FIG. 3 effectively illustrates the presence of velocity differences corresponding to changes in the geologic structure and material properties in different regions between the wells. The tomograms are shown in order proceeding clockwise around the wells beginning with well 12 a. Of course, each tomogram is generated independently in accordance with the method of the prior art. It is noted that the tomogram between wells 12 b and 12 c is repeated in the figures for purposes of illustrating velocity “ties” or continuity of velocity between adjacent tomograms sharing a common well. Ideally, adjacent tomograms should produce velocities that are identical at the common well.
With attention now directed primarily to FIG. 3, three limitations of prior art two dimensional tomography are apparent. First, there is no velocity information is obtained in an area 16, since the prior art process of generating a tomogram is 2D in nature and, therefore, makes no estimate of the velocities in area 16. Second, velocity ties at common wells are quite poor. As an example, an area 20, indicated within a dashed rectangle, exhibits particularly inconsistent velocity between adjacent tomograms sharing well 12 c. Third, the method becomes increasingly approximate as the out of plane geologic structure deviates from pure horizontal layering.
The present invention provides a sweeping solution to all of the foregoing limitations of conventional crosswell seismic imaging and, at the same time, includes other highly advantageous features which have not been seen heretofore in crosswell seismic methods such as, for example, the ability to include data types beyond that of crosswell datasets.
As will be described in more detail hereinafter, a method of characterizing the value of a particular property, for example, seismic velocity, of a subsurface region of ground is disclosed herein. This method, like the method of the prior art, uses seismic data which relates to the property of interest. However, in one aspect of the present invention, the value of the particular property is represented, as determined by the seismic data across the region, using at least one continuous analytic function. In one feature, the continuous analytic function is a polynomial such as, for example, a Chebychev polynomial.
In another aspect of the present invention, the seismic data includes data derived from at least one crosswell dataset for the subsurface region of interest and may include other data. In this instance, the present invention may simultaneously utilize data from a first crosswell dataset in conjunction with the other data. In one feature, the other data may include one or more additional crosswell datasets which are simultaneously used with the initial dataset. In another feature, the other data may include borehole geophysical data, other than crosswell seismic data, which is used simultaneously with the data obtained from the first crosswell dataset.
In still another aspect of the present invention, the first crosswell dataset maybe used simultaneously with the other data in a way which defines the value of the property being characterized in three dimensions within the subsurface region of interest.
In yet another aspect of the present invention, crosswell datasets derived using a borehole or boreholes having a high deviation from the vertical direction or derived using even a generally horizontally extending borehole or boreholes are useful in the method of the present invention. With particular regard to traveltime tomography, it should be appreciated that traveltimes identified between any source and receiver position within the subsurface region of interest are generally considered as being inherently useful irrespective of the fact that different traveltimes may have been produced from different crosswell datasets.
The foregoing aspects are at least in part attributable to the method of the present invention including the formulation of a highly advantageous common earth model. The method provides for establishing a specific numerical value of a particular property at any position within a subterranean region of ground encompassing at least two boreholes from which at least one crosswell seismic dataset is generated relating to the particular property and, in formulation of the model, includes the steps of fitting a set of vertically spaced layer boundaries within and across the region encompassing the boreholes in a predetermined way such that a series of layers is defined between the layer boundaries. Initial values of the particular property are established between the layer boundaries and across the subterranean region using a series of continuous analytic functions, at least one of which corresponds to each one of the series of layers. Thereafter, using the crosswell seismic profile(s) or any other appropriate measured data which is available, the continuous analytic functions are adjusted to more closely match the value of the particular property across the subterranean region of ground such that the value of the particular property is determined for any selected point within the subterranean region. In one feature, the continuous analytic functions are adjusted in an iterative manner until such time that a predetermined stopping condition is reached.
The present invention is also highly effective with regard to representing subsurface regions exhibiting complex geological structure such as, for example, dips, high velocity contrasts and discontinuities including faults.
The present invention may be understood by reference to the following detailed description taken in conjunction with the drawings briefly described below.
FIG. 1 is a diagrammatic perspective view representing one actual subterranean region penetrated by a plurality of boreholes between certain ones of which crosswell datasets were obtained.
FIG. 2 is a diagrammatic twodimensional “unwrapped” planar view of the boreholes shown in FIG. 1, shown here to illustrate the relative depths of the boreholes, including constrictions, and to clearly depict the relationship of prior art twodimensional tomograms taken between the boreholes as laid out in the presentation of FIG. 3.
FIG. 3 is a grayscale representation of a series of prior art twodimensional tomograms illustrating planes of view taken between the boreholes of FIG. 1 in the unwrapped view of FIG. 3, shown here to illustrate poor crossline ties between adjacent tomograms sharing the same borehole and further showing areas of the region where no coverage is provided by the prior art tomograms.
FIG. 4 is a diagrammatic perspective view representing another actual subterranean region penetrated by a plurality of boreholes between certain ones of which crosswell datasets were obtained.
FIG. 5 is a flow diagram illustrating one approach for initially formulating the common earth model of the present invention.
FIG. 6 is a diagrammatic perspective view again illustrating the region initially shown in FIG. 4 including the boreholes therein, shown here to illustrate formation tops identified along the depth of each borehole.
FIG. 7 is a diagrammatic perspective view illustrating the addition of firstorder continuous surfaces extending across the subsurface region and fitted using common picks among the formation tops initially illustrated in FIG. 6.
FIG. 8 is a diagrammatic perspective view illustrating the addition of secondorder continuous surfaces extending across the subsurface region and fitted using common picks among the formation tops initially illustrated in FIG. 6.
FIG. 9 is a diagrammatic perspective view illustrating an initial series of thirdorder continuous surfaces extending across the subsurface region and fitted using common picks among the formation tops initially illustrated in FIG. 6.
FIG. 10 is an elevational diagrammatic view showing two of the boreholes illustrated in FIG. 9 including the initial series of thirdorder surfaces and further illustrates additional, interpolated thirdorder surfaces fitted between the initial series of thirdorder surfaces.
FIG. 11 is a flow diagram illustrating steps included in the overall method of the present invention.
FIG. 12 is an elevational diagrammatic view showing the boreholes illustrated in FIG. 10 including the initial and interpolated surfaces and further illustrating intersection points of a fan of rays with the surfaces wherein the rays extend from a source position in one borehole to a plurality of receiver positions distributed along the depth of the other borehole.
FIG. 13 is an elevational diagrammatic view showing a portion of the boreholes illustrated in FIG. 12 including one exemplary raypath extending between the boreholes and intersecting certain ones of the thirdorder surfaces, shown here to illustrate the coordinates of the raypath at these intersection points.
FIGS. 1417 are diagrammatic elevational views illustrating fans of rays extending between two of the boreholes in an exemplary region including a small slow velocity layer, shown here to illustrate the iterative raybending approach of the method of the present invention and the rapid convergence achieved in the iteration.
FIG. 18 is a diagrammatic elevational view illustrating a fan of rays extending between two boreholes in an actual region including strong velocity contrasts, shown here to illustrate the way in which the fan of rays is forced outside of the interwell plane defined between the boreholes as a result of the velocity contrasts, as a example of only one instance in which the present invention is highly advantageous.
FIG. 19 is a plot of an exemplary residual function illustrating an overall global minimum and local minima.
FIG. 20 is a grayscale representation of a tomogram produced from the application of the threedimensional technique of the present invention, for direct comparison with the prior art tomograms of FIG. 3, wherein the tomogram produced in accordance with the present invention exhibits complete consistency across the extents of the region with no gaps in coverage.
Having previously described FIGS. 13, attention is immediately directed to FIG. 4 which is a perspective view of a subterranean region of ground generally indicated by the reference numeral 30. Region 30 includes five boreholes/wells indicated as 32 ae. It is of interest to note that the borehole geometry of FIG. 4 is representative of an actual geometry found in the Buena Vista Hills oil field of Kern County, Calif. and is often called a ‘fivespot’. In 1997, a crosswell survey was taken in this field yielding four crosswell datasets. For all profiles, well 32 c(653 z) was the source well, and receivers were placed in wells 32 a(554), 32 b(553), 32 d(564) and 32 e(563). It should be appreciated that a prior art tomographic image could be produced for each one of these datasets. In fact, prior art tomograms were prepared for each of these profiles but exhibited unrealistic characteristics due to the inability of prior art methods in dealing with significant deviation from the vertical of well 32 c(653 z). As will be described in detail at appropriate points below, these prior art tomograms exhibited poor crossline ties, and the apparent geologic structure in the tomograms was not structurally consistent with well log correlations and other geologic information from the area. Consequently, the twodimensional tomograms of the prior art were deemed to be unusable by the interpreting geoscientist.
Still referring to FIG. 4, in the Buena Vista field of region 30, wells are most commonly drilled near vertical. This is seen to be the case in the instance of wells 32 a, b, d and e. However, as mentioned, well 32 c is deviated more significantly at approximately 30° from vertical in the interval depicted. It should be appreciated that the amount of deviation exhibited by well 32 c is small compared to other regions (not shown). In these other regions, wells with generally horizontal trajectories through the reservoir zone are very commonly drilled including, for example, multiple highcost wells drilled from offshore platforms. Based on Applicant's experience, deviation of 30° or more from vertical introduces near insurmountable errors in the deviation correction mechanisms employed by the prior art. As will be seen hereinafter, the present invention inherently overcomes the deviation correction problem in a heretofore unseen and highly advantageous way. In fact, results produced by the method of the present invention may actually be enhanced in instances where highly deviated or horizontal wells are encountered, for reasons which will be described.
Having described region 30 and referring now to FIGS. 5 and 6, a basic framework will now be provided regarding the application of the method of the present invention in this region using an associated, highly advantageous model which will be referred to hereinafter as a “common earth” model. FIG. 5 illustrates certain steps in the method of the present invention generally indicated by the reference numeral 40 and beginning with Start step 42. Initially, it should first be noted that information in the form of well logs is normally available for each borehole in the region. These well logs may be produced using various techniques including, but not limited to wireline logging, drilling cuttings logging, core sampling, vertical seismic profiling, and measurement while drilling (MWD). The particular technique employed for producing the well logs is of significance only to the extent that the geologic characteristic of interest such as, for example, velocity can be characterized with reasonable accuracy in the ground immediately surrounding the borehole. Using the characteristic as measured by the well logs, changes indicative of strata boundaries known in the art as “formation tops” are identified in step 44 along the depth of each borehole. In the present example, nine formation tops have been identified for each well. In the instance of well 32 a, the formation tops are indicated by the reference numbers 46 ai. It is to be understood that the method of the present invention is not limited to the use of well logs in the identification of formation tops. That is, any suitable a priori information may be used in accordance with the present invention. Such a priori information may include, but is not limited to information derived from sensing methods such as surface seismic, borehole seismic, electromagnetic methods, gravity and magnetic methods, geologic and geostatistical modelling. Note that method of FIG. 5 may be used to subsequently identify formation tops in a latter portion of the overall method of the present invention wherein the locations of formation tops may be reassessed in view of the iterative process of model development and inversion described herein. That is, step 44 may be modified to incorporate the use of the model to estimate formation top presence rather than the initial use of well logs. Thus, the entire process, including all of the steps of FIG. 5 to be described, may be repeated more than once to further refine the final model using the previous model as the starting point for determining the formation tops in the next iteration.
Following identification of the formation tops, the method proceeds with step 48 in which common horizons are defined across the region. Specifically, formation tops associated with potential strata boundaries (not shown) extending across the region are identified in a known manner. It should be mentioned that, at this point in the method, the common horizon picks identify only potential strata boundaries. While the selected potential boundaries in many cases ultimately turn out to represent actual strata boundaries, the method of the present invention accounts not only for changing the locations and shape of potential boundaries utilized in the common earth model in a flexible manner, but also provides for adding virtually any number of additional boundaries, as will be further described. Further, the identification of potential strata boundaries is not restricted to points on or near to actual wells, and in some cases information to guide the estimation of potential strata boundaries in arbitrary positions in the subsurface may be available such as from one of the sensing methods mentioned above. One example of a potential boundary is represented by formation tops 50 ae. Overall, nine potential common horizons/boundaries are identified, consistent with the number of formation tops identified for each well.
Referring to FIGS. 5 and 7, the method of the present invention continues at step 50 wherein the common horizon picks identifying potential boundaries are used in a highly advantageous way. Specifically, a series of surfaces 52 ac are fitted to the common horizon picks for each potential boundary. Only three surfaces are shown for purposes of clarity, however, it is to be understood that all nine surfaces are actually used, corresponding to the identified potential horizons, and many more could be added as appropriate to the specific area of subterranean ground. It should be appreciated that the mathematical characteristics and formulation of surfaces 52 have been considered carefully in view of the objectives of the present invention. These objectives and the manner in which the common earth model satisfies the objectives will be discussed in detail immediately hereinafter and at appropriate points below.
Continuing with a description of the common earth model of the present invention, one mathematical formulation which has proven to be highly advantageous in defining surfaces 52 utilizes 3rd or higher order Chebychev polynomials. In particular, Chebychev polynomials are used to determine z as a function of x and y for each surface 52 in the form:
where C_{0 }through C_{9 }are the Chebychev coefficients describing a two dimensional surface with cubic structure. It should be noted that these polynomials are characterized using only ten coefficients. In this way, high spatial resolutions are provided with far fewer model parameters than are seen in typical pixelized prior art models, as will be discussed in further detail at an appropriate point below. The ten coefficients can be grouped into four thirdorder terms (C6, C7, C8 and C9), three secondorder terms (C3, C4 and C5), two firstorder terms (C1 and C2) and a constant term (C0). In one feature of the method of the present invention, it is recognized that the weighted distribution of the coefficients in the Chebychev polynomials is highly advantageous as a direct means of controlling the shape or smoothness of the surfaces defined by the functions. By manipulating the weights of the higher order terms (the quadratic and cubic terms involving second and third powers in x and y) the degree of topology in the surfaces may be controlled. In this way, the weighting formulation which is used for the polynomials in a particular geologic setting can be determined based on a priori information such as, for example, the generalized topology established by the predicted horizons/boundaries in view of the formation tops. In this regard, the mathematical complexity of the overall model can be significantly reduced by utilizing the lowest possible order solution which still describes the geology of the region of interest with sufficient accuracy. As general examples of weighted solutions, a firstorder surface is simply a dipping plane that minimizes the “residual” distances between the actual horizon picks and the surface at each well. Surfaces 52, as illustrated in FIG. 7, are firstorder surfaces. Examination of FIG. 7 reveals that, for firstorder surfaces applied to the topology of region 30, the surface may not intersect all of the horizon picks which define it. This is clearly evident for surface 52 c. Therefore, first order surfaces are not considered as appropriate for region 30. Higher order surfaces will be considered immediately hereinafter.
Attention is now directed to FIGS. 8 and 9 which illustrate second and thirdorder surfaces, respectively, applied to the horizon picks of region 30. Initially, it is noted that either secondorder or thirdorder surfaces fit to the horizon picks in region 30 will exactly intersect each well at the depth of the horizon pick. Characteristically, there is enough “freedom” with the addition of the quadratic terms in a secondorder expression to provide good fit for even closely grouped wells with steep structural contours. In this regard, FIG. 8 shows three of nine secondorder surfaces 62 ac fit to the horizon data of FIG. 6 while FIG. 9 shows three of nine third order surfaces indicated as 64 ac. It can be seen in both FIGS. 8 and 9 that the second and thirdorder surfaces intersect the horizon picks of each well. In view of a priori information including well logs or any other suitable data, it may be determined that the secondorder surfaces are most suitable. However, third or higher order surfaces are even more general and can be employed in areas with still greater structural complexity. In either instance, FIGS. 8 and 9 illustrate the threedimensional consistency of the surfaces for any combination of profiles between the wells.
Referring now to FIGS. 5 and 10, it should be appreciated that the sparse number of surfaces defined by the formation tops represent a starting point for the construction of the remainder of the common earth model to be used herein. In step 68, additional surfaces 70 are interpolated between the nine potential surfaces which were identified using the formation tops. FIG. 10 illustrates wells 32 a and 32 c in region 30 and considers the use of third order surfaces 64 ac from FIG. 9. As in previous examples, only three, 64 ac, of the nine potential surfaces identified by formation tops are specifically indicated for purposes of clarity. Thus, the common earth model ultimately includes an overall set 72 of close vertically spaced layers with each layer surface described by a Chebychev polynomial. Spacing between the layers is typically on the order of one to two meters in most typical oil or gas reservoirs. However, the spacing may be varied depending upon the available seismic information and other factors such as, for example, sedimentary heterogeneity, sampling constraints, or frequency content of the seismic data. In the present example, the selected spacing was 25 feet for purposes of display clarity. In this regard, it should be appreciated that vertical resolution in the model representation described herein is directly tied to the vertical spacing between the surfaces. It should also be appreciated that even though FIG. 10 displays only the area between two wells, all of the surfaces are valid, and do not intersect anywhere in the entire 3D model. Any suitable technique may be employed to determine the values of the coefficients for the nine formation top derived polynomials. In the present example, Lanczos Decomposition (SVD) was used to invert the horizon data to establish the coefficients of these polynomials. Thereafter, the coefficients of the interpolated polynomials may be established by any other suitable method of parameter estimation.
It should be appreciated that, in the common earth model of the present invention, using a polynomial characterization of geologic layering is highly advantageous in that the actual geologic structure of a region of interest can in many cases be closely mimicked prior to the inversion process. However, the remarkable flexibility of this model will be brought further to light in the remaining discussions.
With continuing reference to FIGS. 5 and 10, the foregoing descriptions of the common earth model of the present invention have focused on the general topology of the region being modeled. Attention will now be directed to the way in which a geologic characteristic or property of interest is described by the model in a highly advantageous and heretofore unseen way. The characteristic or property of interest may include but is not limited to velocity, attenuation and impedance. In addition, any mathematical formulation using at least one of these or other properties are considered to be within the scope of the present invention so long as teachings disclosed herein are utilized in the formulation. Moreover, combinations of these properties as well as attributes of wave propagation such as, for example, phase, amplitude, and spatial/temporal derivatives thereof and combinations of these attributes are also considered to be within the scope of the present invention. Because the polynomial layer boundaries which make up overall layered arrangement 72 are continuous across region 30, it should be appreciated that a series of continuous layers is defined between these boundaries and across region 30. In FIG. 10, several of these layers are indicated by the reference number 74. For purposes of the common earth model, the properties within each layer are assumed to be homogeneous in the z direction. The characteristic or property of interest is then represented across region 30 for each layer using at least one continuous analytic function. For this purpose, a second series of Chebychev polynomials is used for each interpolated layer defined within the initial series of continuous boundaries to characterize the value of the property in the xy across the extents of region 30, as will be further described. It is submitted that this characterization of a seismic or seismic derived property, in and by itself, is highly advantageous and has not been seen heretofore as applied to borehole seismic imaging.
Still discussing the polynomial characterization of geologic properties of the present invention, one property which has traditionally proven to be useful is seismic velocity. In the present example, the inverse of velocity i.e., slowness has been chosen for representation using Chebychev polynomials since integration of slowness along a raypath gives the traveltime of that path, as will be further described. Representation of the slowness fields as Chebychev polynomials is a key aspect of this formulation for reasons which will become apparent below. For the moment, however, it should be noted that this formulation enables analytical calculation of all of the partial derivatives required for both forward modeling and the tomographic inversion in closed form. Partial derivatives of the velocity field and raypaths with respect to each of the coefficients are a necessary component of most tomographic inversion approaches.
Turning to FIG. 11 and having generally described the common earth model of the present invention, a general discussion of the inversion process is now provided which will serve as an initial framework for considering the application of the common earth model within the method of the present invention and which will further serve in subsequent discussions pointing out the advantages of the present invention. It should also be mentioned that the use of the common earth model will initially be considered using direct arrival traveltime data, as is typically done in conventional twodimensional tomography, and other data types will be considered at appropriate points below. The inversion process shown in FIG. 11 is generally indicated by the reference number 100 and begins with a start step 102 and the associated starting model which is established in a highly advantageous way in step 103 encompassing the steps shown in FIG. 5 and discussed earlier. Thereafter, in forward modeling step 104, a forward modeling step is generally performed in view of actual traveltimes measured between source and receiver positions. Forward modeling of direct arrival traveltimes can be accomplished in a highly advantageous manner to be described. At step 106, the difference between traveltimes calculated during forward modeling step 104 and the actual traveltimes is established in a way which determines a residual function. That is, the residual function sets forth the difference between the forward modeled traveltimes and the actual traveltimes.
Following step 106, an iterative process is initiated in step 108 wherein step 110 is next executed in the event that certain specified termination criteria have not been met. In step 110, a velocity inversion is performed. The inversion optimizes updates to the model parameters with the objective of reducing the residuals. The information used in the inversion is fundamentally some sensitivity function of model parameters based on ray trajectories. Usually, the inversion process finds a vector of model updates. In general, partial derivatives of traveltime with respect to changes in the model parameters calculated during forward modeling step 104 (in this case, the coefficients of the Chebyshev polynomials that describe slowness within each layer are used). There may some optimum scaling that can be applied to this vector to maximize the rate of convergence. The inversion used in the present invention utilizes a linear search to locate the optimum scale factor for the vector of model updates. In step 112, the model parameters are updated.
After the update, the method returns to step 104 repeating the forward modeling step. The overall iterative process in FIG. 11 of inversion, updating of the model parameters, and calculation of traveltimes for the new model, with the objective to minimize the objective function or residuals, in this case traveltime residuals, is continued until at some point a criterion for stopping the iteration is reached. In this instance, if the termination condition is satisfied in step 108, step 114 indicates that the model has been optimized and the method concludes with a stop step 116. Generally, it is not possible to reduce the residuals to zero for various reasons including, but not limited to errors in the actual traveltime data. Therefore, the objective is to seek a “global minimum” for the traveltime residuals representative of a geologically plausible state of model parameters in which modeled traveltimes are as close as possible to the actual data. As the model parameters are updated, traveltime residuals should generally decrease with the iterative process.
With the general steps of the tomographic inversion process of the present invention set forth immediately above, discussions will now be provided with regard to the use of the common earth model of the present invention within the context of these steps in particular consideration of the fact that 3D crosswell imaging has not heretofore been seen. One focus of interest will be that of the specific advantages of the common earth model and how these advantages can be exploited in the instance of 3D traveltime inversion.
Initially considering parameterization of the common earth model, it should be appreciated that two requirements have been met by the present invention for a 3D model configured for simultaneous inversion of multiple profiles at high spatial resolution. Specifically, the common earth model provides sufficient detail for highresolution and is considered to be as welldetermined as possible for the overall inversion process. The term “determined” refers here to the condition of the inversion in view of the ratio of data to unknowns. An underdetermined problem contains more unknowns than can be resolved by the data. Underdetermined inversion results are nonunique and sensitive to small changes in the weighting or constraints used in the inversion. Overdetermined inversions contain more data than unknowns, and results are therefore unique, subject to the constraints imposed by the parameterization of the model. However, not all of the available detail is extracted from the data dependent upon the degree of overdeterminism. In contrast, a welldetermined model extracts the maximum resolution from the data. Historically, the two requirements above have been mutually exclusive in that traditional crosswell models use pixels for 2D. Increasing spatial resolution in pixelized models requires increasing the number of model parameters (pixels), and makes the inverse problem less welldetermined. Thus, the common earth model is considered to be highly advantageous in terms of having achieved an optimum balance between parameters and data.
One of ordinary skill in the art will appreciate that the development of a highly effective three dimensional model in the form of the common earth model presents the need for development of a three dimensional forward modeling technique in step 104 of FIG. 11. To that end, a highly advantageous three dimensional raybending technique will be disclosed which is particularly well suited for use with the common earth model for velocity imaging. Moreover, several key advantages of the common earth model parameterization become clear as this raybending methodology is presented. The subject raybending technique is, in essence, based on the calculation of partial derivatives of traveltime with respect to a raypath. One advantage of the technique resides in rapid convergence even in the instance of extremely heterogeneous models. For example, the raybending technique is capable of determining accurate raypaths for a fastest arriving arrival, which for the purposes of this illustration and the remaining discussion is termed the direct arrival, and thus is capable of determining the traveltime of the same with good accuracy, even in the presence of velocity contrasts that readily cause prior art ray tracing methods to fail.
Attention is now directed to the manner in which the raybending technique operates. Specifically, the technique minimizes the traveltime for a given ray. If the final trajectory satisfies the added constraint of least time (based on Fermat's principle), it can be considered an accurate model of the direct ray path. In the common earth model, traveltime is calculated using line integrals through the velocity fields/layers as given by the second series of Chebychev polynomials described in the foregoing discussions. One assumption of note in the raybending technique is that the boundary spacing (i.e., layer thickness) in the common earth model is sufficiently small such that a raypath is piecewise linear across each layer. As described above, the polynomial formulation actually represents slowness, since integration of slowness yields traveltime. Thus, for a ray segment between two of the model surfaces, the slowness field S(x,y) is represented for each layer with a thirdorder Chebychev polynomial in the form:
where C_{0}C_{9 }are the ten Chebychev coefficients.
The traveltime (T) for the straight ray segment between points A and B on adjacent surfaces can be written:
where dr is an incremental step along the path between A and B. For the straight segments between surfaces, this integral is readily calculated by performing parametric integration. Note that partial spatial derivatives of traveltime with respect to the locations A and B can be advantageously analytically calculated in closed form. Performing the integration yields:
T(x _{1} ,y _{1} ,z _{1} ,x _{2} ,y _{2} ,z _{2} ,C _{0} ,C _{1} ,C _{2} , C _{3} ,C _{4} ,C _{5} ,C _{6} ,C _{7} ,C _{8} ,C _{9})=
where T is traveltime and C_{0}C_{9 }are the ten Chebychev coefficients. It should be appreciated that equations 2 and 4 are relatively simple in form. In contrast, expressions for traveltime and slowness seen in prior art models such as, for example, pixelated models can be extremely complex, requiring relatively enormous computational power merely for the production of twodimensional tomograms. Remarkably, the simplicity of the expressions used in the common earth model of the present invention is attributed to the polynomial formulation of the common earth model. In this regard, it should be appreciated that in the initial stages of the parametric expansion the traveltime expression contained over one hundred terms. However, during derivation of the foregoing expressions, an enormous number of terms cancel, leaving only the terms seen. This highly advantageous result was unanticipated. Moreover, other aspects of the velocity inversion to be described and the forward modeling inversion under discussion are highly advantageous. For example, the matrix to be inverted in either instance has been found to be banddiagonal and symmetric.
Turning to FIG. 12, the raybending technique of the present invention, as used with the common earth model, begins by finding the intersection points of straight rays with the initial series of Chebychev boundary surfaces for source and receiver locations between which a traveltime is available. FIG. 12 illustrates a portion of region 30 including wells 32 a and 32 c. A source position 120 is located along well 32 c while a plurality 122 (only three of which are indicated) of receiver positions are located along well 32 a. Third order Chebychev polynomial surfaces 64 are also shown extending across region 30 in conjunction with a fan of straight rays 124 between source position 120 and each of receiver positions 122 such that the rays form a plurality of intersections 126 with surfaces 64 denoted by a plurality of “crosshairs” at the intersection locations. For thirdorder Chebychev polynomial surfaces, intersections 126 are found by solving the roots of a cubic equation. For one particular ray, if the source position is represented by the coordinates (x_{1},y_{1},z_{1}) and the receiver position is represented by the coordinates (x_{2},y_{2},z_{2}), then a straight line connecting source and receiver can be described by the two equations: x=D_{1}z+D_{2}, and y=D_{3}z+D_{4}. The four coefficients, D_{1}, D_{2}, D_{3}, and D_{4 }can be found by substituting the coordinate values for x and y at the source and receiver positions.
Substituting these expressions for x and y into equation 4 for a selected one of the boundary surfaces 64 yields a cubic equation in z with a real root lying at the intersection of that particular straight ray 124 and the selected surface 64. By solving one cubic for each surface in the model, all of the intersection points are obtained as required for ray tracing/bending with regard to known traveltimes. With all of the intersections to hand, the partial derivatives of traveltime with respect to moving the intersection point may be determined. Because the intersection points are constrained to remain on the model surfaces, the derivatives are calculated only in two dimensions (x and y). Once again, it is noted that, because an analytic expression for traveltime is used, these spatial partial derivatives are advantageously determined in closed form.
Referring to FIG. 13, in order to minimize traveltime, a simultaneous inversion of all of the intersection locations is employed. It is noted that Newton's algorithm performs well in this regard for the common earth model. FIG. 13 diagrammatically illustrates raypath 124 a in region 30 between wells 32 a and 32 c. Raypath 124 a intersects boundary surfaces 64 dg. The intersections of raypath 124 a with surfaces 64 dg are indicated as (x1,y1), (x2,y2), (x3,y3) and (x4,y4), respectively. Newton's method does require the matrix of second partial derivatives of traveltime with respect to x and y of the intersection points, however a more accurate update step could be performed using second derivatives as well as first derivatives. Referring to equation 4 it can be seen that derivatives of x or y can be taken in closed form. These second derivatives can be calculated in closed form at little cost in time or resources. To illustrate the technique, the vector of intersection points that contains pairwise the x and y locations of all the intersection points for a given ray is formulated as:
The vector of first partial derivatives of traveltime with respect to these positions is then
and the matrix of second partial derivatives is:
where T is traveltime. With this notation, the application of Newton's algorithm can be written:
Where T is traveltime given in equation 4, X_{i }and X_{j }are vectors containing the locations of the surface intersection, and λ_{i }is a vector of perturbations to the intersection locations designed to minimize traveltime, so as to define a system of equations. By solving this system of equations, a vector of updates to the model parameters, λ_{i }is found. One method for finding λ_{i }is to perform a linear section search for a scale factor to apply to the model update vector. As mentioned, one highly advantageous feature which arises in using this technique resides in the fact that the matrix of second (Frechet) derivatives has been found to be banddiagonal and symmetric, significantly limiting the number of second derivatives for which values must be determined. A number of other advantages are attendant to this feature including the ability to readily factorize the matrix at far less cost than, for example, using SVD, as will be appreciated by one familiar with SVD.
Referring collectively to FIGS. 1417, the raybending technique introduced immediately above is highly advantageous with regard to its convergence properties. Specifically, in models with velocity contrasts of 25% most rays require fewer than 10 iterations, as will be demonstrated. To illustrate this rapid convergence FIGS. 1417 show raypaths between a source position 118 and a plurality of receiver positions 119 in a model region 120 between two boreholes 122 a and 122 b. The model includes fine layering and a 20% vertical velocity gradient. The velocity field and model layering have not been shown for purposes of clarity. However, it is noted that the uppermost portion of the model contains a small slow velocity layer within a circular area indicated by reference number 124 (FIG. 14). FIG. 14 further illustrates a fan 128 of straight rays between the boreholes. The results of raybending in accordance with the present invention are illustrated after one iteration (FIG. 15) resulting in a fan 130 of rays, two iterations (FIG. 16) resulting in a fan 132 of rays and one hundred iterations (FIG. 17) resulting in a fan 134 of rays. It is evident that the raypaths are optimized by the first few iterations, as seen in FIG. 16. In fact, it has been observed that very little change occurs after only five iterations (not shown). Comparison of FIGS. 16 and 17 reveals that there is little change in the rays from two iterations to onehundred iterations. Thus, the raybending technique of the present invention is highly advantageous with regard to efficiency. Still another advantage of the raybending technique will be presented immediately hereinafter.
Attention is now directed to FIG. 18 serving to illustrate another advantage of the three dimensional raybending technique of the present invention which results from the use of the common earth model. This figure represents an actual region 140 between two deviated wells 142 a and 142 b. Region 140 includes strong velocity contrasts. A fan 144 of rays is shown between the wells after having undergone a sufficient number of iterations. It is observed that the combination of deviated wells and strong velocity contrasts has been found to be capable of forcing a group 146 of rays 144 outside of what, in twodimensional tomography, is conventionally the model space. That is, at least portions of rays 146 are not contained by the interwell plane. It should be recognized in this regard that any model parameterization restricted to an interwell plane would fail in the circumstances shown in FIG. 18. Moreover, the common earth model handles this region with no special consideration, and, indeed, raypaths 146 shown in FIG. 18 do satisfy Fermat's minimum traveltime requirement. In contrast, the prior art typically employs a “shooting” method, rather than raybending. One problem with the shooting approach is assignable to the fact that often the traveltimes for raypaths in areas, for example, of complex structure, as demonstrated by raypaths 146, cannot be obtained. Hence, resolution deteriorates. Generally, prior art ray tracing methods have increasing difficulty as models become more complex, and crucial raypaths, through the regions with the highest velocity contrasts, cannot be obtained. Ironically, these regions of high velocity contrast are typically of great interest to the field engineer or geoscientist. The raybending technique employed in conjunction with the common earth model has not only proven to be effective in regions of complex structure, but also has proven to be very robust and stable, even for models with very high velocity contrasts. By using the raybending approach, accurate raypaths are assured through any geologically reasonable model including, of course, the common earth model. In the case of the common earth model, the raybending approach also offers significant computational efficiency advantages over a rayshooting method.
Referring again to FIG. 11, the model iteration and inversion process of the method of the present invention will now be considered using the common earth model. One general goal in the development of the overall nonlinear model iteration process of the present invention resides in automating the process, particularly the inversion process, to the greatest extent possible. Typically, the starting common earth model would be formulated, for example, by fitting surfaces based on welllog data, picking a dataset from available data for the region of interest such as, for example, direct arrival traveltimes from one or more crosswell data profiles and, thereafter, initiating the model iteration and inversion process until such time that the optimum final result is achieved. Traditionally, this goal has been unmet. Often, prior art inversion algorithms require constant user attention to prevent models from diverging or providing geologically unrealistic solutions. Techniques using shooting ray tracing methods in particular suffer from this problem. For reasons which will be further discussed, conventional methods using pixelized models rarely produce a truly global minimum in traveltime residuals unless particular attention has been paid to inversion constraints during the course of iteration. In contrast, the present invention has markedly, if not fully, achieved this goal through the use of a highly advantageous inversion technique within the overall nonlinear model iteration and inversion process to be described immediately hereinafter.
One significant problem which was encountered in the inversion process used herein regards establishing values of the polynomials used in the common earth model describing the property of interest. Specifically, the value of the property was seen to be well behaved in the areas of the boreholes, however, at some locations away from the boreholes, sometimes unrealistic values (for example, infinity) were seen. Such a result is considered to be unacceptable in accordance with the present invention and geologic plausibility. In order to resolve this problem and to provide for an essentially “automated” inversion process, constraints have been employed in a highly advantageous way which has not been seen in the crosswell seismic imaging arts. Specifically, a continuation constraint approach is utilized. The continuation constraint approach is configured for first resolving longwavelength features. Thereafter, contributions from higher wavenumber features are included. In this manner, spatial resolution and variability in the model is successively increased.
Attention is now directed to FIG. 19 which illustrates an exemplary residual function for a onedimensional or oneparameter problem generally indicated by the reference number 160. The residual function, R, is plotted along a vertical axis 162 of the figure against successive model iterations along a horizontal axis 164. Residual function 160 is illustrated at some arbitrary point in the overall inversion process serving to illustrate characteristics of typical residual functions. In particular, residual function 160 includes a global minimum 166. In this regard, the reader is reminded that the objective of the inversion process of FIG. 11 is to minimize the residual function by converging on global minimum 166. However, the residual function locus also includes local minima 168. In attempting to converge on global minimum 166, it is noted that a “frequency limited” approach has been applied, for example, in prior art pixelated models. However, the approach was applied by first using a coarse pixel grid and, thereafter, adding pixels to the grid. For this and other reasons, prior art pixelated models notoriously converged on local minima such as 168 in the absence of near constant user attention during the inversion. While there are a number of inversion approaches that, in principle, allow one to successively incorporate higher spatial frequencies into a tomographic inversion (e.g. a LevenbergMarquart approach) the prior art does not discuss any specific implementations of an automatic approach to successively incorporate higher wavenumber terms into the inversion process. In view of the prior art, one of ordinary skill in the art would most likely suggest limiting frequency response in the common earth model by beginning the inversion process using only lower order terms of the functions and, thereafter, adding the higher order terms in a manner which is consistent with the prior art pixelized approach. This approach, however, is not considered to be acceptable herein because the truncation or roundoff error introduced by changing the model has undesired effects on convergence. The approach of the present invention, however, operates in an essentially opposite manner, and takes advantage of the polynomial formulation of the model space. Specifically, continuation constraints are applied to the functions with the full variation capability of the functions present. The initial “low frequency” part of the model is then recovered by forcing model variability to be small (forcing the model to be smooth) in the early stages of the inversion process. For a given initial level of constraints forcing model smoothness, it can be determined mathematically when the global minimum has been reached by examining the residual function itself. At this point, the constraints forcing model smoothness are relaxed, allowing the inversion to recover higher frequency components of the model. This process continues until ultimately the constraints forcing model smoothness are negligible and the inversion finds the highest frequency components in the model that are available in the data (the raypaths and arrival times). It has been found that the continuation constraint approach of the present invention almost invariably converges on global minimum 166 in what is considered to be an optimum manner most usually with no user input during the overall inversion process. Traveltime residuals obtained in the continuation constraint approach have typically been two orders of magnitude less than residuals obtained with unconstrained tomography.
In prior art unconstrained tomography, the objective function that is minimized is constructed solely from traveltime errors. With the continuation approach outlined above, it is useful to examine an objective function constructed additionally from the constraints. The function being minimized then has two parts: a part due solely to traveltime misfit, and a part due to how well the current model satisfies the current constraints that force model smoothness. In particular, the objective function is examined in order to determine when to take the next “continuation step” wherein the constraints on the model are reduced. A continuation step is taken when certain termination conditions are met for the current constraints.
The termination conditions for a given fixed value of constraints are determined by examining the objective function that is being minimized: the combination of “data errors” (traveltime residuals) and “penalty errors” (the misfit due to the constraints). There are two termination/stopping conditions that must be met. The first is evidenced by a minimal decrease in the overall objective function. In doing so, successive iterations are compared and the residual itself is examined. When the current iteration produces a residual less than some small fraction of the previous (for example, 0.5%) iteration, the condition is considered as being satisfied. If the previous residual is denoted as “resid_old” and the new or next residual is denoted as “resid_new”, this condition can be expressed in the following form:
where rtol is for example 0.5% or 0.005.
The second stopping condition is concerned with how well the current model parameters and constraints are “solved” in terms of the linear equations. The inversion equations are set up as follows:
where A is the matrix of derivatives of traveltime with respect to the model parameters (the coefficients of the Chebyshev polynomials describing slowness), dp is the vector of updates to the current model parameters (changes in the actual coefficients of the Chebyshev polynomials describing slowness), and dt is the vector of traveltime residuals (the “picked” traveltimes minus the traveltimes computed in the forward modeling step from the Common Earth Model). In addition, the quantities: “AdpN”, the “norm” of the matrixvector product A times dp, and dtN, the “norm” of the righthandside residual vector dt are utilized. The second stopping condition requires the ratio AdpNh/dtN be less than some tolerance (for example, 0.1). The idea being that if the problem Adp=dt is solved, then the quantity Adp is orthogonal to the quantity (dt−Adp). If the problem Adp=dt is solved to a reasonable approximation, then the matrixvector product of the derivative matrix A and the solution vector dp should be close to orthogonal to the difference (dt−Adp) of the residual vector dt minus the matrixvector product Adp. The second stopping condition is expressed as:
AdpN/dtN<Ntol, (9)
where Ntol is, for example, 0.1. When both stopping conditions described above are satisfied, the next “continuation step” decreases the constraints, thereby reducing smoothness to continue the inversion process outlined above.
Having provided an understanding of the common earth model and its method of use, it is now appropriate to discuss associated advantages in view thereof In essence, the present invention is thought to sweep aside prior crosswell tomographic techniques for a number of important reasons. For example, the common earth model and associated method of the present invention are formulated in three dimensions. In and by itself, this advantage has not been seen heretofore and may, in fact, not even fall into full appreciation until such time that the present invention is placed into everyday use. However, certain aspects of this advantage have already become evident and are themselves highly advantageous. For instance, considering traveltime tomography from a single pair of boreholes, the present invention accommodates and accounts for arbitrary well geometries, including horizontal and highly deviated wells. It also provides more accurate representation of actual geologic structure, especially out of plane dip and high velocity contrasts. Further, the present invention provides for the simultaneous use of any number of crosswell seismic data profiles which are available for a region of interest. That is, by executing the full method one time, a single, self consistent, model of the overall region of interest is provided which is capable of providing image data with regard to any portion of the region. Moreover, the problem seen in the prior art and discussed above with regard to poor crossline ties between adjacent profiles or tomograms has been eliminated, as will be seen immediately hereinafter.
With reference to FIG. 3, described previously, and FIG. 20, a direct comparison of results produced in accordance with the present invention will be made to the prior art tomograms of FIG. 3. The latter illustrates an unwrapped view of region 10 which combines a number of prior art tomograms necessarily generated independently between wells 12 ad. FIG. 20 is a grayscale reproduction of an actual tomogram, indicated by the reference number 180, produced using the common earth model and method of the present invention for region 10 of FIG. 1. For purposes of comparison, FIG. 20 illustrates the same unwrapped view as FIG. 1, however, the view of FIG. 20 is taken through the three dimensional space of the common earth model. In comparing these figures, it should immediately be noted that the common earth model provides a completely uniform, consistent image across its entire extents. That is, the problem of poor ties between adjacent profiles has been eliminated. As a specific example, region 20 is indicated for comparison with its appearance in FIG. 3. In FIG. 3, region 20 is representative of poor ties. With regard to the tomogram of FIG. 20, it should be appreciated that the present invention contemplates presentation of images to a user in any of a number of different visualization formats. That is, the common earth model is thought to be readily adaptable to display presentations either presently available or to be developed. For example, “difference tomograms” which display the difference between velocity images generated at two times (for example before and after a steam injection), generated using the common earth model approach discussed above could be represented in three dimensions with areas of large differences clearly visible and distinguishable in extent. Furthermore, the 3D nature of the common earth model lends itself readily to display of three dimensional movies which incorporate the true nature of deviated wells and complex geology.
Continuing with a comparison of FIGS. 3 and 20, the reader may recall that constrictions in certain wells caused the tomograms of FIG. 3 to exhibit areas 16 and 18 in which no coverage is provided by the tomograms. In contrast, the common earth model and method of the present invention, do not produce such gaps using the same datasets as were used in producing the prior art tomograms. That is, tomogram 180 is entirely consistent across its extents including those regions which previously lacked coverage. This advantage is directly attributable to the threedimensional formulation of the common earth model in conjunction with the limited parameterization through the polynomial formulation of the surfaces and properties between the surfaces. In and by itself, this advantage represents an important development in the art, however, in conjunction with the ability to provide a uniform image anywhere in a modeled region, a truly remarkable improvement is provided. Still further advantages remain for discussion.
As noted, the common earth model and associated method of the present invention provide for the simultaneous use of a plurality of crosswell seismic data profiles. In fact, the inherent characteristics of the common earth model advantageously permit the simultaneous use of essentially any suitable geologic or geophysical data in conjunction with one or more crosswell profiles. This advantage is initially best understood by considering traveltime tomography using, for example, direct arrival traveltimes. One of ordinary skill in the art will recognize that some number of direct arrival traveltimes result from a particular crosswell seismic profile. Moreover, each traveltime is defined between a source position and a receiver position. The present invention additionally recognizes that, with the provision of the threedimensional common earth model, there is no need to separately process traveltimes obtained from different profiles. In accordance with the present invention, any traveltime between source and receiver positions in the region of interest is useful, irrespective of which crosswell profile produced that traveltime. Therefore, in a region such as, for example, region 30 of FIG. 4 where a number of crosswell profiles are available, the present invention simultaneously uses all of the direct arrival traveltimes which the profiles yield. Many limitations of prior art tomography are thus eliminated. For example, well deviation is of no concern so long as reasonably accurate source and receiver positions can be located in the region of interest. In fact, highly deviated or completely horizontal wells are considered as advantageous herein since traveltimes produced from such wells result in improved azimuth sampling.
Having initially described the use of direct arrival traveltimes immediately above, it is now appropriate to discuss other types of geologic data which are useful in the present invention. First, it should be recognized by one of ordinary skill in the art that the foregoing model iteration and inversion approach can be used on measured values other than direct arrival traveltime with the stipulation that the raypath is known or previously defined by direct arrival traveltime inversion. Examples of other types of measured values that can be inverted using this approach include amplitude of the direct arrival, center frequency of the direct arrival, bandwidth of the direct arrival and instantaneous phase of the direct arrival. Estimates of earth properties such as attenuation can be obtained by inverting these other measured values. It should also be recognized by one of ordinary skill in the art that raypaths other than the direct arrival between source and receiver can be used in the inversion process. For example, P−S converted, transmitted arrivals could be picked and inverted with this approach. The traveltimes of reflected arrivals from known interfaces could also be used in this approach.
More generally, other types of geologic data can include both seismic and nonseismic data. Electromagnetic data, especially crosswell electromagnetic datasets can be processed by the general method of the common earth model and continuation constraint inversion process described herein, either independently or in conjunction with crosswell seismic data. The crosswell seismic and electromagnetic datasets are differently sensitive to material property variations in the subterranean earth, and can be used jointly to better resolve the imaging of specific earth property fields. In such a case, the common earth model representation of velocity and electrical properties, such as resistivity, could use a single set of common layer boundaries, with each material or data property represented by a separate polynomial within each layer. This is an inherent advantage of the common earth model described herein, in that multiple properties can be assigned and manipulated for each layer simply by establishing a polynomial, or set of polynomials for each property in each layer. For joint seismic and electromagnetic datasets, the model iteration and inversion process of FIG. 11 can be extended to include electromagnetic forward modeling as well as the ray tracing. The inversion process in this case would iterate to minimize a composite residual function corresponding to both electromagnetic data and direct traveltime residual minimization. The model iterations could be computed independently or using a cross coupling of parameters such as seismic velocity variation with resistivity. The specific implementation would depend to a large degree on the confidence with which the characteristic petrophysical properties are known for the specific geological area. However the model representation and inversion method are not restricted in their handling of multiple properties and multiple physics, either independently or in an interlinked form for representation and estimation of geologic structure and material properties. Similarly, additional physics and measurement methods are not restricted to seismic and electromagnetic, and all other practical data types, including but not limited to gravity, magnetic, flow, tracer and pressure transient data as well as geological analogue maybe accommodated within the model and method either as measured and forward modeled data or as constraints or a priori information.
Further, in areas where there is a significant lateral discontinuity in geologic structure, such as at a fault, where there may be an abrupt change in property variation and even possible continuation in another layer, the model representation can be adjusted by using two, or more, polynomial function representations for the value of a given property within a layer, such that a linear combination of two or more polynomial functions determines the value of the property parameter. In the case of a fault like discontinuity, each parameter representative polynomial is each modulated by a further function which is essentially scaled linearly between zero and one throughout the xy space of the layer such that the linear combination of the two or more polynomial functions determines the value of the parameter on either side of each lateral interface within a model layer.
It has been mentioned previously how other types of data can be used to estimate formation tops, especially other types of seismic data. In a specific refinement of the method, the resulting velocity, or other parameter field, output by the process at the end of the inversion and model iteration process could be used to establish a better starting model and the process could then repeated. For example in the case of a velocity model, the velocity field output by the process could be spatially differentiated to provide estimates of the actual positions of the interfaces between geologic strata which in most cases would be more accurate than the initial interpolations between formation tops estimated at wells or other arbitrary points in space. Accordingly, the layer boundary surfaces could be constructed in a much more geologically conformal manner than would be possible initially without the benefit of the results of the model iteration and inversion. The results of the second model so developed could be reapplied to develop a third, further refined model, and so on. In each case the distribution of the geological or material properties determined by the model could be used to update the layer boundary surfaces for the next version of the model, so ultimately, both the layer surfaces and the properties distribution within the layers converge to better depict the actual distribution of geologic strata and material properties of the subsurface. In such multiple iterations of the overall model process, the use of property information to update the layer boundaries is not restricted to a single property; multiple properties could be used conjunctly, or sequentially in different iterations.
The common earth model introduced above is wellsuited also to crosswell reflection imaging. Generally, the first step in reflection imaging is to generate accurate velocity information for the survey area. The invention described above functions in this regard and can be considered a critical step in achieving practical 3D reflectance imaging from crosswell data. Following the construction of a velocity model from the tomography process, the same common earth model and raybending framework described above can be used for reflection raytracing. Reflection raytracing generates “trajectories” for a set of subsurface locations of reflection points associated with a single sourcereceiver pair and all of the model surfaces illustrated, for example, in FIG. 10. That is, the location on each of the model surfaces of the reflection point is found that minimizes the arrival time for the overall raypath through the current velocity model. One significant improvement of this invention compared to the prior art 2D models resides in the fact that reflections due to dipping reflectors will be accurately positioned in 3D. In prior art crosswell reflection raytracing methods, the assumption of no outofplane dip forces all reflection events to be located in the plane of the wells. However, for stratigraphic beds having geologic structure with outofplane dips, the true reflection location may be significantly away from the plane of the wells.
With reflection trajectories calculated through an accurate velocity model, it is possible to “map” time to depth. In this process, reflected seismic signals generated for a known combination of source and receiver geometry can be located accurately in the 3D model. The datadependant required information for mapping is the sourcereceiver geometry and the arrival times of the reflected energy. The reflection trajectories described above, together with the reflection times calculated with the trajectories, illustrate the path of reflection energy. Finally, for a given sourcereceiver pair and signal time, we can locate from the trajectories, velocity model and reflection times the subsurface location giving rise to the reflection. The results of reflectance imaging could be used as a basis for updating the layer boundary surface starting model for another iteration of the process as described above.
Two other particular aspects of other seismic data used bear special mention. High resolution processing of surface seismic, such as acoustic impedance inversion can be used to estimate properties and strata boundaries on a subwavelength scale. The results of this type of processing can be incorporated into the method to provided detailed estimations of formation tops, especially where well data is not available, and also as an initial starting model and constraints for properties inversion. Another form of seismic data of special interest is that of single well seismic imaging. This is a particular extension of crosswell seismic in which the source and receiver are located the same borehole. Single well seismic imaging is a relatively immature method and poses significant challenges in data acquisition. However the imaging objectives are generally similar to that of conventional crosswell seismic, and typically velocity tomography and reflectance imaging methods are applied. Conventional processing is generally limited for single well seismic imaging because conventional vertical well, planar geometry representation are of little practical meaning when applied to single well dataset geometries. However, the common earth model representation is entirely valid and appropriate to single well seismic imaging, as is the raybending method for velocity tomography.
As noted above, principal prior art methods utilize iterative backprojection through a slowness model, and therefore are subject to the constraints and limitations of the crosswell aperture. In contrast, the present invention utilizes approximations to a layered earth through polynomial surfaces and velocity fields so as to circumvent these limitations, providing much higher vertical resolution than could be practically expected out of any of the conventional traveltime tomography techniques, especially when applied to realistic three dimensional earth models. Therefore, the present invention is considered to be completely set apart and distinct from any of the principal prior art techniques and provides significant resolution and computational advantages when compared to any presently known form of crosswell traveltime tomography.
Moreover, the approach of the present invention which serves to gain resolution while obtaining a multiwell consistent 3D result relies on parameterizing surfaces and the velocity fields defined as layers between these surfaces as Chebychev polynomials. This affords highly effective means for computing Frechet derivatives and for forward modeling through the earth model using a raybending method. In contrast, Guiziou, et al, (1996) use a parameterization of a velocity field having a linear gradient between surfaces and fits the lateral variation in velocity with Bsplines. In fact, the actual surfaces used by Guiziou are arbitrary and are not polynomials. Guiziou uses this parameterization because it deals with tomographic inversion of reflection data obtained from a surface seismic geometry. Were the Guiziou approach used herein for crosswell traveltime tomography, the functionality achieved herein could not be realized, because such surfaces would not accommodate the use of a raybending technique and the Frechet derivatives would be far more complex to compute. In addition, there is no justification for assigning arbitary surfaces, as described by Guiziou, because in crosswell imaging information is typically to hand with regard to the surfaces concentrated at certain points, such as at the wells, whereas, in principle, Guiziou has information about the surfaces distributed everywhere throughout his 3D model. Therefore, because the common earth model parameterizes the surfaces as well as the velocity field, and applies to crosswell rather than surface seismic data, it is distinct, and offers far reaching advantages over the method proposed by Guiziou.
With brief regard to the prior art approach taken by Chiu and Stewart, 1987. The parameterization of surfaces is discussed as either polynomials or as sinusoidal functions, however this art does not teach the parameterization of the velocity field as a firstorder or higher polynomial. Rather, Chiu and Stewart treat the velocity field between the surfaces as constant. Parameterization of the common earth model in such a manner would be inconsistent with incorporation of realistic lateral heterogeneity in a computationally efficient way.
Thus, in view of the foregoing, the present invention is set apart from the prior art in its use of (1) Chebychev polynomial representation of both surfaces and layer parameters in general traveltime tomography, (2) raybending computation and analytic derivative computation in conjunction with Chebychev representation in general traveltime tomography and (3) use of polynomial surfaces and polynomial property fields in crosswell direct arrival traveltime tomography and imaging. These distinctions have beneficial implications in the implementation of crosswell seismic imaging including traveltime tomography and further extend to the application to reflectance imaging and other advanced methods, as described above.
It should be appreciated that the concepts of the present invention, as used in the method taught herein and in the highly advantageous common earth model, may be applied in a number of different ways by one of skill in the art. Therefore, the present examples and methods are considered as illustrative and not restrictive, and the invention is not to be limited to the details given herein, but may be modified within the scope of the appended claims.
Claims (92)
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

US09152935 US6388947B1 (en)  19980914  19980914  Multicrosswell profile 3D imaging and method 
Applications Claiming Priority (1)
Application Number  Priority Date  Filing Date  Title 

US09152935 US6388947B1 (en)  19980914  19980914  Multicrosswell profile 3D imaging and method 
Publications (1)
Publication Number  Publication Date 

US6388947B1 true US6388947B1 (en)  20020514 
Family
ID=22545073
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

US09152935 Active US6388947B1 (en)  19980914  19980914  Multicrosswell profile 3D imaging and method 
Country Status (1)
Country  Link 

US (1)  US6388947B1 (en) 
Cited By (86)
Publication number  Priority date  Publication date  Assignee  Title 

US20020042678A1 (en) *  20000807  20020411  Dimitri Bevc  Typing picks to horizons in migration velocity analysis 
US20020050989A1 (en) *  20000908  20020502  Sanstrom William C.  System and method for attaching drilling information to threedimensional visualizations of earth models 
US20030105440A1 (en) *  20000106  20030605  UniCharm Corporation  Waterdecomposable absorbent article 
US20030132934A1 (en) *  20011212  20030717  Technoguide As  Three dimensional geological model construction 
US6662109B2 (en) *  20010419  20031209  Institut Francais Du Petrole  Method of constraining by dynamic production data a fine model representative of the distribution in the reservoir of a physical quantity characteristic of the subsoil structure 
US20030236645A1 (en) *  20020126  20031225  Blaz Santic  Evaluating method for testing of simple geometry and complete gauging of the work pieces with defined tolerance zones 
US20040010373A1 (en) *  20020711  20040115  Smits Jan W.  Intelligent diagnosis of environmental influence on well logs with modelbased inversion 
US20040068376A1 (en) *  20021004  20040408  Baker Hughes Incorporated  Walkaway tomographic monitoring 
US6778907B1 (en) *  20000712  20040817  Core Laboratories Lp  Method for estimation of propagation paths for seismic signals 
GB2401453A (en) *  20030508  20041110  Schlumberger Holdings  Method of determining the petrophysical properties of a borehole and the region surrounding it 
US20050114032A1 (en) *  20031121  20050526  Xishuo Wang  Seismic data interpolation system 
US20050117452A1 (en) *  20031201  20050602  Calvert Rodney W.  Welltowell tomography 
US6917564B2 (en) *  20000727  20050712  Westerngeco, L.L.C.  Method of processing seismic data 
US20060047429A1 (en) *  20040824  20060302  Adams Steven L  Method of estimating geological formation depths by converting interpreted seismic horizons from the time domain to the depth domain 
US20060149520A1 (en) *  20030221  20060706  Mickaele Le RavalecDupin  Method for more rapidly producing the representative stochastic model of a heterogeneous underground reservoir defined by uncertain static and dynamic data 
US20060195264A1 (en) *  20050225  20060831  Galil El Askary Said A  Method and apparatus for estimating distance to or from a geological target while drilling or logging 
US20060203877A1 (en) *  20050310  20060914  Heyman Joseph S  Dynamic acoustic thermometer 
US20070021951A1 (en) *  20030723  20070125  Lee Seislink Wook B  3D velocity modeling, with calibration and trend fitting using geostatistical techniques, particularly advantageous for curved for curvedray prestack time migration and for such migration followed by prestack depth migration 
US7248539B2 (en) *  20030410  20070724  Schlumberger Technology Corporation  Extrema classification 
US20080270028A1 (en) *  20070427  20081030  Schlumberger Technology Corporation  Pixel based inversion method for surface electromagnetic measurement 
US20090164188A1 (en) *  20071221  20090625  Tarek Habashy  Method for upscaling a reservoir model using deep reading measurements 
US20090164187A1 (en) *  20071221  20090625  Tarek Habashy  Method for reservoir characterization and monitoring including deep reading quad combo measurements 
US7644765B2 (en)  20061020  20100112  Shell Oil Company  Heating tar sands formations while controlling pressure 
US20100018718A1 (en) *  20060928  20100128  Krebs Jerome R  Iterative inversion of data from simultaneous geophysical sources 
US7673786B2 (en)  20060421  20100309  Shell Oil Company  Welding shield for coupling heaters 
US20100118655A1 (en) *  20081108  20100513  Andres Chavarria  Progressive 3D vertical seismic profiling method 
US20100228527A1 (en) *  20090305  20100909  Schlumberger Technology Corporation  Coarsening and splitting techniques 
US20100225647A1 (en) *  20090305  20100909  Schlumberger Technology Corporation  Right sizing reservoir models 
US7798221B2 (en)  20000424  20100921  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US7798220B2 (en)  20070420  20100921  Shell Oil Company  In situ heat treatment of a tar sands formation after drive process treatment 
US7831134B2 (en)  20050422  20101109  Shell Oil Company  Grouped exposed metal heaters 
US7866386B2 (en)  20071019  20110111  Shell Oil Company  In situ oxidation of subsurface formations 
US20110066404A1 (en) *  20090916  20110317  Chevron U.S.A. Inc.  Method for creating a 3d rock representation using petrophysical data 
US20110087459A1 (en) *  20091009  20110414  Zazovsky Alexander F  Cleanup prediction and monitoring 
US7942203B2 (en)  20030424  20110517  Shell Oil Company  Thermal processes for subsurface formations 
US20110120724A1 (en) *  20080811  20110526  Krohn Christine E  Estimation of Soil Properties Using Waveforms of Seismic Surface Waves 
US20110131020A1 (en) *  20090909  20110602  Conocophillips Company  Dip guided full waveform inversion 
US20110153300A1 (en) *  20081106  20110623  Holl James E  System and Method For Planning A Drilling Operation 
US20110194379A1 (en) *  20100210  20110811  Sunwoong Lee  Methods For Subsurface Parameter Estimation In Full Wavefield Inversion And ReverseTime Migration 
US20110238390A1 (en) *  20100329  20110929  Krebs Jerome R  Full Wavefield Inversion Using Time Varying Filters 
US8064287B2 (en) *  20061228  20111122  Rock Solid Images, Inc.  Method for interpreting seismic data and controlled source electromagnetic data to estimate subsurface reservoir properties 
US8107317B2 (en)  20061228  20120131  Schlumberger Technology Corporation  Technique and system for performing a cross well survey 
US8151880B2 (en)  20051024  20120410  Shell Oil Company  Methods of making transportation fuel 
US8151907B2 (en)  20080418  20120410  Shell Oil Company  Dual motor systems and nonrotating sensors for use in developing wellbores in subsurface formations 
WO2012047323A1 (en) *  20100930  20120412  Chevron U.S.A. Inc.  System and method for declustering subsurface data on a perslice basis 
US8224163B2 (en)  20021024  20120717  Shell Oil Company  Variable frequency temperature limited heaters 
US8220539B2 (en)  20081013  20120717  Shell Oil Company  Controlling hydrogen pressure in selfregulating nuclear reactors used to treat a subsurface formation 
US8327932B2 (en)  20090410  20121211  Shell Oil Company  Recovering energy from a subsurface formation 
US8355623B2 (en)  20040423  20130115  Shell Oil Company  Temperature limited heaters with high power factors 
WO2013036241A1 (en)  20110909  20130314  Landmark Graphics Corporation  Crosswell seismic surveying in a deviated borehole 
WO2013015879A3 (en) *  20110728  20130321  Chevron U.S.A. Inc.  System and method for perturbing an initial horizonpicking solution to follow local features of a volume 
WO2013052566A1 (en) *  20111007  20130411  Shell Oil Company  Using dielectric properties of an insulated conductor in a subsurface formation to assess properties of the insulated conductor 
US8437998B2 (en)  20100927  20130507  Exxonmobil Upstream Research Company  Hybrid method for full waveform inversion using simultaneous and sequential source method 
CN103415786A (en) *  20110608  20131127  雪佛龙美国公司  System and method for seismic data inversion by nonlinear model update 
US8608249B2 (en)  20010424  20131217  Shell Oil Company  In situ thermal processing of an oil shale formation 
US8627887B2 (en)  20011024  20140114  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US8631866B2 (en)  20100409  20140121  Shell Oil Company  Leak detection in circulated fluid systems for heating subsurface formations 
US8660824B2 (en)  20101223  20140225  Schlumberger Technology Corporation  Discretely gridding complex fracturing for simulation 
US8688381B2 (en)  20101201  20140401  Exxonmobil Upstream Research Company  Simultaneous source inversion for marine streamer data with crosscorrelation objective function 
US8694299B2 (en)  20100507  20140408  Exxonmobil Upstream Research Company  Artifact reduction in iterative inversion of geophysical data 
US8701769B2 (en)  20100409  20140422  Shell Oil Company  Methods for treating hydrocarbon formations based on geology 
US8756042B2 (en)  20100519  20140617  Exxonmobile Upstream Research Company  Method and system for checkpointing during simulations 
US8767508B2 (en)  20100818  20140701  Exxonmobil Upstream Research Company  Using seismic P and S arrivals to determine shallow velocity structure 
US8775143B2 (en)  20100927  20140708  Exxonmobil Upstream Research Company  Simultaneous source encoding and source separation as a practical solution for full wavefield inversion 
US8812282B2 (en)  20080321  20140819  Exxonmobil Upstream Research Company  Efficient method for inversion of geophysical data 
US8820406B2 (en)  20100409  20140902  Shell Oil Company  Electrodes for electrical current flow heating of subsurface formations with conductive material in wellbore 
US8892413B2 (en)  20110330  20141118  Exxonmobil Upstream Research Company  Convergence rate of full wavefield inversion using spectral shaping 
US8917199B2 (en)  20110413  20141223  Raytheon Company  Subterranean image generating device and associated method 
US8931580B2 (en)  20100203  20150113  Exxonmobil Upstream Research Company  Method for using dynamic target region for well path/drill center optimization 
US8990053B2 (en)  20110331  20150324  Exxonmobil Upstream Research Company  Method of wavelet estimation and multiple prediction in full wavefield inversion 
US9016370B2 (en)  20110408  20150428  Shell Oil Company  Partial solution mining of hydrocarbon containing layers prior to in situ heat treatment 
US9033042B2 (en)  20100409  20150519  Shell Oil Company  Forming bitumen barriers in subsurface hydrocarbon formations 
US20150177413A1 (en) *  20131220  20150625  Schlumberger Technology Corporation  Method And Apparatus To Generate A Crosswell Data Set 
KR101552261B1 (en)  20140901  20150909  한국지질자원연구원  Method and systme for modeling unconformity surface of strata in elastic wave prospecting signal 
US9140812B2 (en)  20110902  20150922  Exxonmobil Upstream Research Company  Using projection onto convex sets to constrain fullwavefield inversion 
US9176930B2 (en)  20111129  20151103  Exxonmobil Upstream Research Company  Methods for approximating hessian times vector operation in full wavefield inversion 
US20160086079A1 (en) *  20140602  20160324  Westerngeco Llc  Properties link for simultaneous joint inversion 
US9309755B2 (en)  20111007  20160412  Shell Oil Company  Thermal expansion accommodation for circulated fluid systems used to heat subsurface formations 
KR101614135B1 (en)  20140901  20160421  한국지질자원연구원  Method for verifying modelled reflecting interface using elastic wave and system therof 
US9595129B2 (en)  20120508  20170314  Exxonmobil Upstream Research Company  Canvas control for 3D data volume processing 
US9605524B2 (en)  20120123  20170328  Genie Ip B.V.  Heater pattern for in situ thermal processing of a subsurface hydrocarbon containing formation 
US9612355B2 (en)  20100709  20170404  Halliburton Energy Services, Inc.  Imaging and sensing of subterranean reservoirs 
US9702993B2 (en)  20130524  20170711  Exxonmobil Upstream Research Company  Multiparameter inversion through offset dependent elastic FWI 
US9702998B2 (en)  20130708  20170711  Exxonmobil Upstream Research Company  Fullwavefield inversion of primaries and multiples in marine environment 
US9772413B2 (en)  20130823  20170926  Exxonmobil Upstream Research Company  Simultaneous sourcing during both seismic acquisition and seismic inversion 
US9910189B2 (en)  20150313  20180306  Exxonmobil Upstream Research Company  Method for fast line search in frequency domain FWI 
Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

US5144590A (en)  19910808  19920901  B P America, Inc.  Bed continuity detection and analysis using crosswell seismic data 
US5481501A (en) *  19930219  19960102  Exxon Production Research Company  Method for simulating crosswell seismic data 
US5583826A (en) *  19920529  19961210  Compagnie Generale De Geophysique  Method of processing signals for geophysical prospecting, the method making use of an improved wave field extrapolation operator 
US5586082A (en) *  19950302  19961217  The Trustees Of Columbia University In The City Of New York  Method for identifying subsurface fluid migration and drainage pathways in and among oil and gas reservoirs using 3D and 4D seismic imaging 
US5691958A (en) *  19950413  19971125  Exxon Production Research Company  Method for determining formation properties from seismic attributes 
Patent Citations (5)
Publication number  Priority date  Publication date  Assignee  Title 

US5144590A (en)  19910808  19920901  B P America, Inc.  Bed continuity detection and analysis using crosswell seismic data 
US5583826A (en) *  19920529  19961210  Compagnie Generale De Geophysique  Method of processing signals for geophysical prospecting, the method making use of an improved wave field extrapolation operator 
US5481501A (en) *  19930219  19960102  Exxon Production Research Company  Method for simulating crosswell seismic data 
US5586082A (en) *  19950302  19961217  The Trustees Of Columbia University In The City Of New York  Method for identifying subsurface fluid migration and drainage pathways in and among oil and gas reservoirs using 3D and 4D seismic imaging 
US5691958A (en) *  19950413  19971125  Exxon Production Research Company  Method for determining formation properties from seismic attributes 
NonPatent Citations (8)
Title 

A. Gersztenkorn and J. A. Scales, Smoothing seismic tomograms with alphatrimmed means, 1988, Geophysical Journal, 92. 
Jean Luc Guiziou, Jean Laurent Mallet, and Raul Madariaga, 3D seismic reflection tomography on top of the COCAD depth modeler, Sep. 1996, Geophysics, vol. 61, No. 5. 
John A. Scales, Paul Docherty, and Adam Gersztenkorn, Regularizaton of nonlinear inverse problems: imaging the nearsurface weathering layer, 1990, Inverse Problems, 6. 
John A. Scales, Tomographic inversion via the conjugate gradient method, Feb. 1987, Geophysics, vol. 52, No. 2. 
Keith A. Meyerholtz, Gary L. Pavlis, and Sally A. Szpakowski, Convolution quelling in seismic tomography, May 1989, Geophysics, vol. 54, No. 5. 
Kris A. Dines and R. Jeffrey Lytle, Computerized Geophysical Tomography, Jul. 1979, Proceedings of the IEEE, vol. 67, No. 7. 
Stephen K.L. Chiu and Robert R. Stewart, Tomographic determination of threedimensional seismic velocity structure using well logs, vertical seismic profiles, and surface seismic data, Aug. 1987, Geophysics, vol. 52, No. 8. 
W. Scott Phillips and Michael C. Fehler, Traveltime Tomography: A comparison of popular methods, Oct. 1991, Geophysics, vol. 56, No. 10. 
Cited By (218)
Publication number  Priority date  Publication date  Assignee  Title 

US20030105440A1 (en) *  20000106  20030605  UniCharm Corporation  Waterdecomposable absorbent article 
US8485252B2 (en)  20000424  20130716  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US8225866B2 (en)  20000424  20120724  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US7798221B2 (en)  20000424  20100921  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US8789586B2 (en)  20000424  20140729  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US6778907B1 (en) *  20000712  20040817  Core Laboratories Lp  Method for estimation of propagation paths for seismic signals 
US6917564B2 (en) *  20000727  20050712  Westerngeco, L.L.C.  Method of processing seismic data 
US6687618B2 (en) *  20000807  20040203  3D Geo Development, Inc.  Typing picks to horizons in migration velocity analysis 
US20020042678A1 (en) *  20000807  20020411  Dimitri Bevc  Typing picks to horizons in migration velocity analysis 
US20020050989A1 (en) *  20000908  20020502  Sanstrom William C.  System and method for attaching drilling information to threedimensional visualizations of earth models 
US6801197B2 (en) *  20000908  20041005  Landmark Graphics Corporation  System and method for attaching drilling information to threedimensional visualizations of earth models 
US6662109B2 (en) *  20010419  20031209  Institut Francais Du Petrole  Method of constraining by dynamic production data a fine model representative of the distribution in the reservoir of a physical quantity characteristic of the subsoil structure 
US8608249B2 (en)  20010424  20131217  Shell Oil Company  In situ thermal processing of an oil shale formation 
US8627887B2 (en)  20011024  20140114  Shell Oil Company  In situ recovery from a hydrocarbon containing formation 
US7542037B2 (en)  20011212  20090602  Schlumberger Technology Corporation  Three dimensional geological model construction 
US7248259B2 (en) *  20011212  20070724  Technoguide As  Three dimensional geological model construction 
US20030132934A1 (en) *  20011212  20030717  Technoguide As  Three dimensional geological model construction 
US20060197759A1 (en) *  20011212  20060907  Technoguide As  Three dimensional geological model construction 
US20030236645A1 (en) *  20020126  20031225  Blaz Santic  Evaluating method for testing of simple geometry and complete gauging of the work pieces with defined tolerance zones 
US7069174B2 (en) *  20020127  20060627  Blaz Santic  Evaluating method for testing of simple geometry and complete gauging of the work pieces with defined tolerance zones 
US20040010373A1 (en) *  20020711  20040115  Smits Jan W.  Intelligent diagnosis of environmental influence on well logs with modelbased inversion 
US6832159B2 (en) *  20020711  20041214  Schlumberger Technology Corporation  Intelligent diagnosis of environmental influence on well logs with modelbased inversion 
US6894949B2 (en) *  20021004  20050517  Baker Hughes Incorporated  Walkaway tomographic monitoring 
US20040068376A1 (en) *  20021004  20040408  Baker Hughes Incorporated  Walkaway tomographic monitoring 
US8224163B2 (en)  20021024  20120717  Shell Oil Company  Variable frequency temperature limited heaters 
US8238730B2 (en)  20021024  20120807  Shell Oil Company  High voltage temperature limited heaters 
US8224164B2 (en)  20021024  20120717  Shell Oil Company  Insulated conductor temperature limited heaters 
US7392166B2 (en) *  20030221  20080624  Institut Francais Du Petrole  Method for more rapidly producing the representative stochastic model of a heterogeneous underground reservoir defined by uncertain static and dynamic data 
US20060149520A1 (en) *  20030221  20060706  Mickaele Le RavalecDupin  Method for more rapidly producing the representative stochastic model of a heterogeneous underground reservoir defined by uncertain static and dynamic data 
US7248539B2 (en) *  20030410  20070724  Schlumberger Technology Corporation  Extrema classification 
US7942203B2 (en)  20030424  20110517  Shell Oil Company  Thermal processes for subsurface formations 
US8579031B2 (en)  20030424  20131112  Shell Oil Company  Thermal processes for subsurface formations 
US20040225441A1 (en) *  20030508  20041111  Schlumberger Technology Corporation, A Texas Corporation  Automated borehole geology and petrophysics interpretation using image logs 
GB2401453B (en) *  20030508  20051221  Schlumberger Holdings  Automated borehole geology and petrophysics interpretation using image logs 
US7133779B2 (en)  20030508  20061107  Schlumberger Technology Corporation  Automated borehole geology and petrophysics interpretation using image logs 
GB2401453A (en) *  20030508  20041110  Schlumberger Holdings  Method of determining the petrophysical properties of a borehole and the region surrounding it 
US20070021951A1 (en) *  20030723  20070125  Lee Seislink Wook B  3D velocity modeling, with calibration and trend fitting using geostatistical techniques, particularly advantageous for curved for curvedray prestack time migration and for such migration followed by prestack depth migration 
US7493241B2 (en) *  20030723  20090217  Lee Wook B  3D velocity modeling, with calibration and trend fitting using geostatistical techniques, particularly advantageous for curved for curvedray prestack time migration and for such migration followed by prestack depth migration 
US7027929B2 (en)  20031121  20060411  GeoX Systems Ltd.  Seismic data interpolation system 
US20050114032A1 (en) *  20031121  20050526  Xishuo Wang  Seismic data interpolation system 
US20050117452A1 (en) *  20031201  20050602  Calvert Rodney W.  Welltowell tomography 
US7046581B2 (en) *  20031201  20060516  Shell Oil Company  Welltowell tomography 
US8355623B2 (en)  20040423  20130115  Shell Oil Company  Temperature limited heaters with high power factors 
US20070118293A1 (en) *  20040824  20070524  Adams Steven L  Method of estimating geological formation depths by converting interpreted seismic horizons the time domain to the depth domain 
US20060047429A1 (en) *  20040824  20060302  Adams Steven L  Method of estimating geological formation depths by converting interpreted seismic horizons from the time domain to the depth domain 
US7433786B2 (en)  20040824  20081007  Adams Steven L  Method of estimating geological formation depths by converting interpreted seismic horizons the time domain to the depth domain 
US7209834B2 (en) *  20050225  20070424  Halliburton Energy Services, Inc.  Method and apparatus for estimating distance to or from a geological target while drilling or logging 
US20060195264A1 (en) *  20050225  20060831  Galil El Askary Said A  Method and apparatus for estimating distance to or from a geological target while drilling or logging 
US20060203877A1 (en) *  20050310  20060914  Heyman Joseph S  Dynamic acoustic thermometer 
WO2006099563A3 (en) *  20050310  20071122  Luna Innovations Inc  Dynamic acoustic thermometer 
WO2006099563A2 (en) *  20050310  20060921  Luna Innovations, Inc.  Dynamic acoustic thermometer 
US7404671B2 (en) *  20050310  20080729  Luna Innovations Incorporated  Dynamic acoustic thermometer 
US7942197B2 (en)  20050422  20110517  Shell Oil Company  Methods and systems for producing fluid from an in situ conversion process 
US8027571B2 (en)  20050422  20110927  Shell Oil Company  In situ conversion process systems utilizing wellbores in at least two regions of a formation 
US8224165B2 (en)  20050422  20120717  Shell Oil Company  Temperature limited heater utilizing nonferromagnetic conductor 
US7860377B2 (en)  20050422  20101228  Shell Oil Company  Subsurface connection methods for subsurface heaters 
US8233782B2 (en)  20050422  20120731  Shell Oil Company  Grouped exposed metal heaters 
US7831134B2 (en)  20050422  20101109  Shell Oil Company  Grouped exposed metal heaters 
US8070840B2 (en)  20050422  20111206  Shell Oil Company  Treatment of gas from an in situ conversion process 
US7986869B2 (en)  20050422  20110726  Shell Oil Company  Varying properties along lengths of temperature limited heaters 
US8230927B2 (en)  20050422  20120731  Shell Oil Company  Methods and systems for producing fluid from an in situ conversion process 
US8606091B2 (en)  20051024  20131210  Shell Oil Company  Subsurface heaters with low sulfidation rates 
US8151880B2 (en)  20051024  20120410  Shell Oil Company  Methods of making transportation fuel 
US7785427B2 (en)  20060421  20100831  Shell Oil Company  High strength alloys 
US8857506B2 (en)  20060421  20141014  Shell Oil Company  Alternate energy source usage methods for in situ heat treatment processes 
US7912358B2 (en)  20060421  20110322  Shell Oil Company  Alternate energy source usage for in situ heat treatment processes 
US7673786B2 (en)  20060421  20100309  Shell Oil Company  Welding shield for coupling heaters 
US8083813B2 (en)  20060421  20111227  Shell Oil Company  Methods of producing transportation fuel 
US7866385B2 (en)  20060421  20110111  Shell Oil Company  Power systems utilizing the heat of produced formation fluid 
US7683296B2 (en)  20060421  20100323  Shell Oil Company  Adjusting alloy compositions for selected properties in temperature limited heaters 
US7793722B2 (en)  20060421  20100914  Shell Oil Company  Nonferromagnetic overburden casing 
US20100018718A1 (en) *  20060928  20100128  Krebs Jerome R  Iterative inversion of data from simultaneous geophysical sources 
US8428925B2 (en)  20060928  20130423  Exxonmobil Upstream Research Company  Iterative inversion of data from simultaneous geophysical sources 
US9495487B2 (en)  20060928  20161115  Exxonmobil Upstream Research Company  Iterative inversion of data from simultaneous geophysical sources 
US8121823B2 (en)  20060928  20120221  Exxonmobil Upstream Research Company  Iterative inversion of data from simultaneous geophysical sources 
US7730946B2 (en)  20061020  20100608  Shell Oil Company  Treating tar sands formations with dolomite 
US7681647B2 (en)  20061020  20100323  Shell Oil Company  Method of producing drive fluid in situ in tar sands formations 
US7677310B2 (en)  20061020  20100316  Shell Oil Company  Creating and maintaining a gas cap in tar sands formations 
US7677314B2 (en)  20061020  20100316  Shell Oil Company  Method of condensing vaporized water in situ to treat tar sands formations 
US7841401B2 (en)  20061020  20101130  Shell Oil Company  Gas injection to inhibit migration during an in situ heat treatment process 
US7673681B2 (en)  20061020  20100309  Shell Oil Company  Treating tar sands formations with karsted zones 
US8555971B2 (en)  20061020  20131015  Shell Oil Company  Treating tar sands formations with dolomite 
US7730947B2 (en)  20061020  20100608  Shell Oil Company  Creating fluid injectivity in tar sands formations 
US7730945B2 (en)  20061020  20100608  Shell Oil Company  Using geothermal energy to heat a portion of a formation for an in situ heat treatment process 
US7845411B2 (en)  20061020  20101207  Shell Oil Company  In situ heat treatment process utilizing a closed loop heating system 
US8191630B2 (en)  20061020  20120605  Shell Oil Company  Creating fluid injectivity in tar sands formations 
US7717171B2 (en)  20061020  20100518  Shell Oil Company  Moving hydrocarbons through portions of tar sands formations with a fluid 
US7644765B2 (en)  20061020  20100112  Shell Oil Company  Heating tar sands formations while controlling pressure 
US7703513B2 (en)  20061020  20100427  Shell Oil Company  Wax barrier for use with in situ processes for treating formations 
US8107317B2 (en)  20061228  20120131  Schlumberger Technology Corporation  Technique and system for performing a cross well survey 
US9310505B2 (en)  20061228  20160412  Schlumberger Technology Corporation  Technique and system for performing a cross well survey 
US8064287B2 (en) *  20061228  20111122  Rock Solid Images, Inc.  Method for interpreting seismic data and controlled source electromagnetic data to estimate subsurface reservoir properties 
US7841425B2 (en)  20070420  20101130  Shell Oil Company  Drilling subsurface wellbores with cutting structures 
US7849922B2 (en)  20070420  20101214  Shell Oil Company  In situ recovery from residually heated sections in a hydrocarbon containing formation 
US7931086B2 (en)  20070420  20110426  Shell Oil Company  Heating systems for heating subsurface formations 
US8327681B2 (en)  20070420  20121211  Shell Oil Company  Wellbore manufacturing processes for in situ heat treatment processes 
US7841408B2 (en)  20070420  20101130  Shell Oil Company  In situ heat treatment from multiple layers of a tar sands formation 
US8459359B2 (en)  20070420  20130611  Shell Oil Company  Treating nahcolite containing formations and saline zones 
US7832484B2 (en)  20070420  20101116  Shell Oil Company  Molten salt as a heat transfer fluid for heating a subsurface formation 
US7798220B2 (en)  20070420  20100921  Shell Oil Company  In situ heat treatment of a tar sands formation after drive process treatment 
US8662175B2 (en)  20070420  20140304  Shell Oil Company  Varying properties of in situ heat treatment of a tar sands formation based on assessed viscosities 
US8791396B2 (en)  20070420  20140729  Shell Oil Company  Floating insulated conductors for heating subsurface formations 
US9181780B2 (en)  20070420  20151110  Shell Oil Company  Controlling and assessing pressure conditions during treatment of tar sands formations 
US8042610B2 (en)  20070420  20111025  Shell Oil Company  Parallel heater system for subsurface formations 
US8381815B2 (en)  20070420  20130226  Shell Oil Company  Production from multiple zones of a tar sands formation 
US7950453B2 (en)  20070420  20110531  Shell Oil Company  Downhole burner systems and methods for heating subsurface formations 
US7640110B2 (en) *  20070427  20091229  Schlumberger Technology Corporation  Pixel based inversion method for surface electromagnetic measurement 
US20080270028A1 (en) *  20070427  20081030  Schlumberger Technology Corporation  Pixel based inversion method for surface electromagnetic measurement 
US8536497B2 (en)  20071019  20130917  Shell Oil Company  Methods for forming long subsurface heaters 
US7866388B2 (en)  20071019  20110111  Shell Oil Company  High temperature methods for forming oxidizer fuel 
US8162059B2 (en)  20071019  20120424  Shell Oil Company  Induction heaters used to heat subsurface formations 
US8272455B2 (en)  20071019  20120925  Shell Oil Company  Methods for forming wellbores in heated formations 
US7866386B2 (en)  20071019  20110111  Shell Oil Company  In situ oxidation of subsurface formations 
US8196658B2 (en)  20071019  20120612  Shell Oil Company  Irregular spacing of heat sources for treating hydrocarbon containing formations 
US8011451B2 (en)  20071019  20110906  Shell Oil Company  Ranging methods for developing wellbores in subsurface formations 
US8146661B2 (en)  20071019  20120403  Shell Oil Company  Cryogenic treatment of gas 
US8146669B2 (en)  20071019  20120403  Shell Oil Company  Multistep heater deployment in a subsurface formation 
US8240774B2 (en)  20071019  20120814  Shell Oil Company  Solution mining and in situ treatment of nahcolite beds 
US8113272B2 (en)  20071019  20120214  Shell Oil Company  Threephase heaters with common overburden sections for heating subsurface formations 
US8276661B2 (en)  20071019  20121002  Shell Oil Company  Heating subsurface formations by oxidizing fuel on a fuel carrier 
US8738341B2 (en) *  20071221  20140527  Schlumberger Technology Corporation  Method for reservoir characterization and monitoring including deep reading quad combo measurements 
US20090164187A1 (en) *  20071221  20090625  Tarek Habashy  Method for reservoir characterization and monitoring including deep reading quad combo measurements 
US20090164188A1 (en) *  20071221  20090625  Tarek Habashy  Method for upscaling a reservoir model using deep reading measurements 
US8498848B2 (en)  20071221  20130730  Schlumberger Technology Corporation  Method for upscaling a reservoir model using deep reading measurements 
US8744817B2 (en)  20071221  20140603  Schlumberger Technology Corporation  Method for upscaling a reservoir model using deep reading measurements 
US8812282B2 (en)  20080321  20140819  Exxonmobil Upstream Research Company  Efficient method for inversion of geophysical data 
US8172335B2 (en)  20080418  20120508  Shell Oil Company  Electrical current flow between tunnels for use in heating subsurface hydrocarbon containing formations 
US8177305B2 (en)  20080418  20120515  Shell Oil Company  Heater connections in mines and tunnels for use in treating subsurface hydrocarbon containing formations 
US8162405B2 (en)  20080418  20120424  Shell Oil Company  Using tunnels for treating subsurface hydrocarbon containing formations 
US9528322B2 (en)  20080418  20161227  Shell Oil Company  Dual motor systems and nonrotating sensors for use in developing wellbores in subsurface formations 
US8636323B2 (en)  20080418  20140128  Shell Oil Company  Mines and tunnels for use in treating subsurface hydrocarbon containing formations 
US8151907B2 (en)  20080418  20120410  Shell Oil Company  Dual motor systems and nonrotating sensors for use in developing wellbores in subsurface formations 
US8752904B2 (en)  20080418  20140617  Shell Oil Company  Heated fluid flow in mines and tunnels used in heating subsurface hydrocarbon containing formations 
US8562078B2 (en)  20080418  20131022  Shell Oil Company  Hydrocarbon production from mines and tunnels used in treating subsurface hydrocarbon containing formations 
US8892410B2 (en)  20080811  20141118  Exxonmobil Upstream Research Company  Estimation of soil properties using waveforms of seismic surface waves 
US20110120724A1 (en) *  20080811  20110526  Krohn Christine E  Estimation of Soil Properties Using Waveforms of Seismic Surface Waves 
US8256512B2 (en)  20081013  20120904  Shell Oil Company  Movable heaters for treating subsurface hydrocarbon containing formations 
US8881806B2 (en)  20081013  20141111  Shell Oil Company  Systems and methods for treating a subsurface formation with electrical conductors 
US8267185B2 (en)  20081013  20120918  Shell Oil Company  Circulated heated transfer fluid systems used to treat a subsurface formation 
US8353347B2 (en)  20081013  20130115  Shell Oil Company  Deployment of insulated conductors for treating subsurface formations 
US9022118B2 (en)  20081013  20150505  Shell Oil Company  Double insulated heaters for treating subsurface formations 
US8220539B2 (en)  20081013  20120717  Shell Oil Company  Controlling hydrogen pressure in selfregulating nuclear reactors used to treat a subsurface formation 
US9129728B2 (en)  20081013  20150908  Shell Oil Company  Systems and methods of forming subsurface wellbores 
US8267170B2 (en)  20081013  20120918  Shell Oil Company  Offset barrier wells in subsurface formations 
US8261832B2 (en)  20081013  20120911  Shell Oil Company  Heating subsurface formations with fluids 
US8281861B2 (en)  20081013  20121009  Shell Oil Company  Circulated heated transfer fluid heating of subsurface hydrocarbon formations 
US9051829B2 (en)  20081013  20150609  Shell Oil Company  Perforated electrical conductors for treating subsurface formations 
US8849640B2 (en)  20081106  20140930  Exxonmobil Upstream Research Company  System and method for planning a drilling operation 
US20110153300A1 (en) *  20081106  20110623  Holl James E  System and Method For Planning A Drilling Operation 
US20140257707A1 (en) *  20081108  20140911  Sr2020, Inc.  Progressive 3D Vertical Seismic Profiling Method 
US20100118655A1 (en) *  20081108  20100513  Andres Chavarria  Progressive 3D vertical seismic profiling method 
US8339396B2 (en)  20090305  20121225  Schlumberger Technology Corporation  Coarsening and splitting techniques 
US20100228527A1 (en) *  20090305  20100909  Schlumberger Technology Corporation  Coarsening and splitting techniques 
US20100225647A1 (en) *  20090305  20100909  Schlumberger Technology Corporation  Right sizing reservoir models 
US8350851B2 (en)  20090305  20130108  Schlumberger Technology Corporation  Right sizing reservoir models 
US8851170B2 (en)  20090410  20141007  Shell Oil Company  Heater assisted fluid treatment of a subsurface formation 
US8434555B2 (en)  20090410  20130507  Shell Oil Company  Irregular pattern treatment of a subsurface formation 
US8327932B2 (en)  20090410  20121211  Shell Oil Company  Recovering energy from a subsurface formation 
US8448707B2 (en)  20090410  20130528  Shell Oil Company  Nonconducting heater casings 
US20110131020A1 (en) *  20090909  20110602  Conocophillips Company  Dip guided full waveform inversion 
US20110066404A1 (en) *  20090916  20110317  Chevron U.S.A. Inc.  Method for creating a 3d rock representation using petrophysical data 
US8818778B2 (en) *  20090916  20140826  Chevron U.S.A. Inc.  Method for creating a 3D rock representation using petrophysical data 
US9121263B2 (en) *  20091009  20150901  Schlumberger Technology Corporation  Cleanup prediction and monitoring 
US20110087459A1 (en) *  20091009  20110414  Zazovsky Alexander F  Cleanup prediction and monitoring 
US8931580B2 (en)  20100203  20150113  Exxonmobil Upstream Research Company  Method for using dynamic target region for well path/drill center optimization 
US20110194379A1 (en) *  20100210  20110811  Sunwoong Lee  Methods For Subsurface Parameter Estimation In Full Wavefield Inversion And ReverseTime Migration 
US8537638B2 (en)  20100210  20130917  Exxonmobil Upstream Research Company  Methods for subsurface parameter estimation in full wavefield inversion and reversetime migration 
US20110238390A1 (en) *  20100329  20110929  Krebs Jerome R  Full Wavefield Inversion Using Time Varying Filters 
US8223587B2 (en)  20100329  20120717  Exxonmobil Upstream Research Company  Full wavefield inversion using time varying filters 
US9399905B2 (en)  20100409  20160726  Shell Oil Company  Leak detection in circulated fluid systems for heating subsurface formations 
US8701769B2 (en)  20100409  20140422  Shell Oil Company  Methods for treating hydrocarbon formations based on geology 
US9022109B2 (en)  20100409  20150505  Shell Oil Company  Leak detection in circulated fluid systems for heating subsurface formations 
US8820406B2 (en)  20100409  20140902  Shell Oil Company  Electrodes for electrical current flow heating of subsurface formations with conductive material in wellbore 
US8701768B2 (en)  20100409  20140422  Shell Oil Company  Methods for treating hydrocarbon formations 
US8833453B2 (en)  20100409  20140916  Shell Oil Company  Electrodes for electrical current flow heating of subsurface formations with tapered copper thickness 
US8631866B2 (en)  20100409  20140121  Shell Oil Company  Leak detection in circulated fluid systems for heating subsurface formations 
US8739874B2 (en)  20100409  20140603  Shell Oil Company  Methods for heating with slots in hydrocarbon formations 
US9033042B2 (en)  20100409  20150519  Shell Oil Company  Forming bitumen barriers in subsurface hydrocarbon formations 
US9127538B2 (en)  20100409  20150908  Shell Oil Company  Methodologies for treatment of hydrocarbon formations using staged pyrolyzation 
US9127523B2 (en)  20100409  20150908  Shell Oil Company  Barrier methods for use in subsurface hydrocarbon formations 
US8880384B2 (en)  20100507  20141104  Exxonmobil Upstream Research Company  Artifact reduction in iterative inversion of geophysical data 
US8694299B2 (en)  20100507  20140408  Exxonmobil Upstream Research Company  Artifact reduction in iterative inversion of geophysical data 
US8756042B2 (en)  20100519  20140617  Exxonmobile Upstream Research Company  Method and system for checkpointing during simulations 
US9612355B2 (en)  20100709  20170404  Halliburton Energy Services, Inc.  Imaging and sensing of subterranean reservoirs 
US8767508B2 (en)  20100818  20140701  Exxonmobil Upstream Research Company  Using seismic P and S arrivals to determine shallow velocity structure 
US8775143B2 (en)  20100927  20140708  Exxonmobil Upstream Research Company  Simultaneous source encoding and source separation as a practical solution for full wavefield inversion 
US8437998B2 (en)  20100927  20130507  Exxonmobil Upstream Research Company  Hybrid method for full waveform inversion using simultaneous and sequential source method 
WO2012047323A1 (en) *  20100930  20120412  Chevron U.S.A. Inc.  System and method for declustering subsurface data on a perslice basis 
US8688381B2 (en)  20101201  20140401  Exxonmobil Upstream Research Company  Simultaneous source inversion for marine streamer data with crosscorrelation objective function 
US8660824B2 (en)  20101223  20140225  Schlumberger Technology Corporation  Discretely gridding complex fracturing for simulation 
US8892413B2 (en)  20110330  20141118  Exxonmobil Upstream Research Company  Convergence rate of full wavefield inversion using spectral shaping 
US9081115B2 (en)  20110330  20150714  Exxonmobil Upstream Research Company  Convergence rate of full wavefield inversion using spectral shaping 
US8990053B2 (en)  20110331  20150324  Exxonmobil Upstream Research Company  Method of wavelet estimation and multiple prediction in full wavefield inversion 
US9016370B2 (en)  20110408  20150428  Shell Oil Company  Partial solution mining of hydrocarbon containing layers prior to in situ heat treatment 
US8917199B2 (en)  20110413  20141223  Raytheon Company  Subterranean image generating device and associated method 
EP2718746A4 (en) *  20110608  20160113  Chevron Usa Inc  System and method for seismic data inversion by nonlinear model update 
CN103415786A (en) *  20110608  20131127  雪佛龙美国公司  System and method for seismic data inversion by nonlinear model update 
WO2013015879A3 (en) *  20110728  20130321  Chevron U.S.A. Inc.  System and method for perturbing an initial horizonpicking solution to follow local features of a volume 
US9140812B2 (en)  20110902  20150922  Exxonmobil Upstream Research Company  Using projection onto convex sets to constrain fullwavefield inversion 
CN103782199A (en) *  20110909  20140507  界标制图有限公司  Crosswell seismic surveying in a deviated borehole 
WO2013036241A1 (en)  20110909  20130314  Landmark Graphics Corporation  Crosswell seismic surveying in a deviated borehole 
WO2013052566A1 (en) *  20111007  20130411  Shell Oil Company  Using dielectric properties of an insulated conductor in a subsurface formation to assess properties of the insulated conductor 
CN104011327B (en) *  20111007  20161214  国际壳牌研究有限公司  Using subterranean formation dielectric properties of the insulated wire to determine the performance of the insulated wire 
CN104011327A (en) *  20111007  20140827  国际壳牌研究有限公司  Using dielectric properties of insulated contductor in subsurface formation to assess properties of insulated conductor 
US9080917B2 (en)  20111007  20150714  Shell Oil Company  System and methods for using dielectric properties of an insulated conductor in a subsurface formation to assess properties of the insulated conductor 
US9309755B2 (en)  20111007  20160412  Shell Oil Company  Thermal expansion accommodation for circulated fluid systems used to heat subsurface formations 
US9176930B2 (en)  20111129  20151103  Exxonmobil Upstream Research Company  Methods for approximating hessian times vector operation in full wavefield inversion 
US9605524B2 (en)  20120123  20170328  Genie Ip B.V.  Heater pattern for in situ thermal processing of a subsurface hydrocarbon containing formation 
US9595129B2 (en)  20120508  20170314  Exxonmobil Upstream Research Company  Canvas control for 3D data volume processing 
US9702993B2 (en)  20130524  20170711  Exxonmobil Upstream Research Company  Multiparameter inversion through offset dependent elastic FWI 
US9702998B2 (en)  20130708  20170711  Exxonmobil Upstream Research Company  Fullwavefield inversion of primaries and multiples in marine environment 
US9772413B2 (en)  20130823  20170926  Exxonmobil Upstream Research Company  Simultaneous sourcing during both seismic acquisition and seismic inversion 
US20150177413A1 (en) *  20131220  20150625  Schlumberger Technology Corporation  Method And Apparatus To Generate A Crosswell Data Set 
US20160086079A1 (en) *  20140602  20160324  Westerngeco Llc  Properties link for simultaneous joint inversion 
US9852373B2 (en) *  20140602  20171226  Westerngeco L.L.C.  Properties link for simultaneous joint inversion 
KR101614135B1 (en)  20140901  20160421  한국지질자원연구원  Method for verifying modelled reflecting interface using elastic wave and system therof 
KR101552261B1 (en)  20140901  20150909  한국지질자원연구원  Method and systme for modeling unconformity surface of strata in elastic wave prospecting signal 
US9910189B2 (en)  20150313  20180306  Exxonmobil Upstream Research Company  Method for fast line search in frequency domain FWI 
Similar Documents
Publication  Publication Date  Title 

Stoffa et al.  Splitstep Fourier migration  
Hicks et al.  Reflection waveform inversion using local descent methods: Estimating attenuation and velocity over a gassand depositWaveform Inversion over a Gas Sand  
US5838634A (en)  Method of generating 3D geologic models incorporating geologic and geophysical constraints  
US5784334A (en)  Method and system for detecting hydrocarbon reservoirs using amplitude versus offset analysis of seismic signals  
US5742560A (en)  Seismic imaging using wave equation extrapolation  
Shipp et al.  Twodimensional full wavefield inversion of wideaperture marine seismic streamer data  
Billette et al.  Velocity macromodel estimation from seismic reflection data by stereotomography  
Shaw et al.  Waveform inversion of seismic refraction data and applications to young Pacific crust  
US6128580A (en)  Convertedwave processing in manylayered anisotropic media  
Michelini et al.  Seismological studies at Parkfield. I. Simultaneous inversion for velocity structure and hypocenters using cubic Bsplines parameterization  
US7355923B2 (en)  Seismic analysis using postimaging seismic anisotropy corrections  
Zelt  Modelling strategies and model assessment for wideangle seismic traveltime data  
US6778909B1 (en)  Seismic data processing systems and methods  
Rawlinson et al.  Seismic traveltime tomography of the crust and lithosphere  
US20100088035A1 (en)  Pseudoanalytical method for the solution of wave equations  
US20060098529A1 (en)  Method for data regulariization for shot domain processing  
US6901333B2 (en)  Method and device for the generation and application of anisotropic elastic parameters  
US5640368A (en)  Migration velocity analysis using limitedaperture and monte carlo migration  
US5570321A (en)  Seismic velocity model optimization method using simulated annearling to determine prestack traveltimes  
US4839869A (en)  Methods for processing converted wave seismic data  
Chauris et al.  Migration velocity analysis from locally coherent events in 2D laterally heterogeneous media, Part I: Theoretical aspects  
US4415999A (en)  Method of confirming seismic data interpretation  
US20100177595A1 (en)  Using Seismic Attributes for Data Alignment and Seismic Inversion In Joint PP/PS Seismic Analysis  
Rickett et al.  Offset and angledomain common imagepoint gathers for shotprofile migration  
Süss et al.  P wave seismic velocity structure derived from sonic logs and industry reflection data in the Los Angeles basin, California 
Legal Events
Date  Code  Title  Description 

AS  Assignment 
Owner name: TOMOSEIS, INC., TEXAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:WASHBOURNE, JOHN K.;RECTOR, JAMES W. III;BUBE, KENNETH P.;REEL/FRAME:009643/0669;SIGNING DATES FROM 19980921 TO 19980924 

AS  Assignment 
Owner name: ENERGY, U.S. DEPARTMENT OF, CALIFORNIA Free format text: CONFIRMATORY LICENSE;ASSIGNOR:TOMOSEIS, INC.;REEL/FRAME:011848/0167 Effective date: 20010302 

AS  Assignment 
Owner name: CORE LABORATORIES LP, TEXAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:TOMESEIS, INC.;REEL/FRAME:013767/0233 Effective date: 20030627 

FPAY  Fee payment 
Year of fee payment: 4 

AS  Assignment 
Owner name: ZSEIS CORPORATION, TEXAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:CORE LABORATORIES LP;REEL/FRAME:022043/0131 Effective date: 20081230 

FPAY  Fee payment 
Year of fee payment: 8 

FPAY  Fee payment 
Year of fee payment: 12 